The scientific pretensions of atheism

The enclosed anthology of articles (mainly from Commentary) by mathematician David Berlinski, summarizes the substantial scientific deficiencies in current attempts to explain the origin of the universe (cosmic evolution), the origin of life (chemical evolution) and the development of life (biological evolution). Few readers are aware of the grave objections to these theories that attempt to explain the marvels of life via the combination of unguided naturalistic processes and chance without reference to intelligent agency.

The blurb to Belinski book The Devil's Delusion: Atheism and Its Scientific Pretensions (Crown Forum, 2008) states: Militant atheism is on the rise. Richard Dawkins, Sam Harris, Daniel Dennett, and Christopher Hitchens have dominated bestseller lists with books denigrating religious belief as dangerous foolishness. And these authors are merely the leading edge of a far larger movement–one that now includes much of the scientific community. “The attack on traditional religious thought,” writes David Berlinski in The Devil’s Delusion, “marks the consolidation in our time of science as the single system of belief in which rational men and women might place their faith, and if not their faith, then certainly their devotion.” A secular Jew, Berlinski nonetheless delivers a biting defense of religious thought. An acclaimed author who has spent his career writing about mathematics and the sciences, he turns the scientific community’s cherished skepticism back on itself, daring to ask and answer some rather embarrassing questions:

Has anyone provided a proof of God’s inexistence? Not even close.

Has quantum cosmology explained the emergence of the universe or why it is here? Not even close.

Have the sciences explained why our universe seems to be fine-tuned to allow for the existence of life? Not even close.

Are physicists and biologists willing to believe in anything so long as it is not religious thought? Close enough.

Has rationalism in moral thought provided us with an understanding of what is good, what is right, and what is moral? Not close enough.

Has secularism in the terrible twentieth century been a force for good? Not even close to being close.

Is there a narrow and oppressive orthodoxy of thought and opinion within the sciences? Close enough.

Does anything in the sciences or in their philosophy justify the claim that religious belief is irrational? Not even ballpark.

Is scientific atheism a frivolous exercise in intellectual contempt? Dead on.

Berlinski does not dismiss the achievements of western science. The great physical theories, he observes, are among the treasures of the human race. But they do nothing to answer the questions that religion asks, and they fail to offer a coherent description of the cosmos or the methods by which it might be investigated.

Commentary February 1998

Was There a Big Bang?

David Berlinski

SCIENCE IS a congeries of great quests, and cosmology is the grandest of the great quests. Taking as its province the universe as a whole, cosmology addresses the old, the ineradicable questions about space and time, nature and destiny. It is not a subject for the tame or the timid.

For the first half of the 20th century, cosmology remained a discipline apart, as a clutch of talented but otherworldly physicists peeped inconclusively at a universe they could barely see: Albert Einstein, of course; the Dutch mathematician Willem de Sitter; the Belgian abbot Georges Lemaître; and the extraordinary Russian mathematician and meteor-ologist Aleksandr Friedmann, destined to die young, or so the story has it, from exposure to the elements suffered while soaring above Moscow in a weather balloon.

When in 1917 Einstein published his first estimates of the size and shape of the universe, telescopes could not penetrate the heavens beyond the Milky Way. Like a sailor endeavoring to measure the depth of the sea from the shore, astronomers lacked the means to probe the heavens further or to probe them in detail.

This has now changed. Information pours from the night skies, terrestrial telescopes hissing and clicking as they rotate to survey distant galaxies. Somewhere in space, the realigned Hubble telescope peers into the unpolluted depths. Physicists have pictures of the backside of beyond, and they appear to have overheard the cosmic cackle that accompanied the very crack of time, as nothingness gave way to light. The cosmologists have come into their own, handling the universe with an easy familiarity and writing book after book in which they explain in exuberant detail how the great things were done.

The Great Cause WHATEVER THE dreams we dream, the existence of the universe has always seemed a riddle beyond reason, if only because our imagination is forever suspended between ideas of creation and timelessness. Many ancient myths depict the universe as the effect of some Great Cause. In the Babylonian epic Enuma Elish, existence is attributed to congress between "primordial Aspu, the Begetter" and "Mummu-Tiamet, she who bore them all." Although the connection cosmic and sexual energies is both familiar and disturbing, it is not congress between gods that is crucial to the myth but the idea that the universe came into being as the result of a Great Cause. And this idea is conveyed by the opening of Genesis as well: "In the beginning, God created the heavens and the earth."

These are words that express an authentically universal concept, one familiar to every culture. But while the concept of creation is common, it is also incoherent. "I venture to ask," the 3rd-century Chinese sage Kuo Hsiang ventured to ask, "whether the Creator is or is not. If He is not, how can He create things? And if He is, then (being one of these things) He is incapable (without self-creation) of creating the mass of bodily forms." That this argument is simple is no reason to think it wrong.

If not creation by a Great Cause, what, then, of timelessness: a universe proceeding sedately from everlasting to everlasting? In Maori myth, the world parents who bring the cosmos into being arise themselves from po, a kind of antecedent gruel or stuff, so that the universe appears as an episode in an infinitely extended drama. Some variant of this idea is also universal, a place in every culture where the weary mind takes refuge. Yet if timelessness offers an escape from the paradoxes of creation, the escape can easily seem an evasion. An everlasting universe is itself an object requiring explanation: if it is unprofitable to ask when it arose, one nonetheless wonders why the damn thing should be there at all.

It is the remarkable claim of contemporary cosmology to have broken, and broken decisively, the restless movement of the mind as it passes from theory to theory and from myth to myth. "Incomplete though it may be," one physicist has written, "the scope of modern scientific understanding of the cosmos is truly dazzling." This is not hyperbole. It is an assessment widely shared among physicists, and thus the standard by which their claims must be judged.

The universe, cosmologists now affirm, came into existence as the expression of an explosion, the cosmos shaking itself into existence from the bang of an initial singularity. It is tempting to think of the event in humanly comprehensible terms--a gigantic explosion or a stupendous eruption, as if, popcorn in hand, we were watching the show from far away. But this is absurd. The Big Bang was not an event taking place at a time or in a place. Space and time were themselves created by the Big Bang, the measure along with the measured.

As far as most physicists are concerned, the Big Bang is now a part of the structure of serene indubitability created by modern physics, an event undeniable as the volcanic explosion at Krakatoa. From time to time, it is true, the astrophysical journals report the failure of observation to confirm the grand design. It hardly matters. The physicists have not only persuaded themselves of the merits of Big Bang cosmology, they have persuaded everyone else as well. The Big Bang has come to signify virtually a universal creed; men and women who know nothing of cosmology are convinced that the rumble of creation lies within reach of their collective memory.

The Cosmic Archeologist

LOOKING AT a few shards of pottery on the desert floor, the archeologist is capable of conjuring up the hanging gardens of the past, the smell of myrrh and honey in the air. His is an act of intellectual reconstruction, one made poignant by the fact that the civilization from which the artifacts spring lies forever beyond the reach of anything but remembrance and the imagination. Cosmology on the grand scale is another form of archeology; the history of the cosmos reveals itself in layers, like the strata of an ancient city.

The world of human artifacts makes sense against the assumption of a continuous human culture. The universe is something else: an old, eerie place with no continuous culture available to enable us to make sense of what we see. It is the hypothesis that the universe is expanding that has given cosmologists a unique degree of confidence as they climb down the cliffs of time.

A universe that is expanding is a universe with a clear path into the past. If things are now far apart, they must at one point have been close together; and if things were once close together, they must at one point have been hotter than they are now, the contraction of space acting to compress its constituents like a vise, and so increase their energy. The retreat into the past ends at an initial singularity, a state in which material particles are at no distance from each other and the temperature, density, and curvature of the universe are infinite.

The cosmic archeologist may now be observed crawling back up the cliffs of time he so recently descended. During the first 10-43 seconds after the Big Bang--10-43 is one over a ten followed by 42 zeros--both matter and radiation fill the void. A reign of fluid interchange obtains, with particles of matter and antimatter exchanging identities. As the primitive goo of the cosmos--what the physicist George Gamow called ylem, the primordial stuff--continues to expand, it continues to cool. Neutrinos, photons, electrons, positrons, neutrons, and protons agitate themselves throughout space. With the temperature dropping, the neutrinos decamp for parts unknown.

At roughly one-and-one-half seconds after the Big Bang, protons and neutrons lose the ability to exchange identities, and the ratio of neutrons to protons in the universe freezes itself at one to six. Three-and-one-half seconds later, the equilibrium between electrons and positrons collapses, and the positrons follow the neutrinos into the void. Three minutes pass. The era of nucleosynthesis begins thereafter. Those neutrons that during the freeze-out found themselves bound to the world's vagrant protons now take up an identity as a form of helium. Other elements wait patiently for the stars to be born so that they may be cooked in their interiors.

The universe continues to expand, pulse, glow, throb, and moan for 400,000 years more, passing insensibly from a place where radiation predominates to an arena where matter has taken charge and is in command. The temperature is now 4,000 degrees Kelvin. The great era of recombination is at hand, a burst of cosmic creativity recorded in the walls of time. Free electrons and protons form hydrogen. The interaction between matter and radiation changes dramatically.

Until recombination, photons found themselves trapped within a cosmic pinball machine, ricocheting from one free charged particle to another, the cosmos frustratingly opaque because frustratingly dense. But hydrogen binds the cosmic debris, and for the first time, light streams from one side of creation to the other. The early universe fills with low- temperature blackbody radiation, the stuff destined to be observed fifteen billion years later in Princeton, New Jersey, as cosmic background radiation (CBR).

The separation of light and matter allows the galaxies to form, gravity binding the drifting dust in space. At last, the universe fills with matter, the stars settling into the sky, the far-flung suns radiating energy, the galaxies spreading themselves throughout the heavens. On the earth that has been newly made, living things shamble out of the warm oceans, the cosmic archeologist himself finally clambering over the lip of time to survey the scene and take notes on all that has occurred.

Such is the standard version of hot Big Bang cosmology--"hot" in contrast to scenarios in which the universe is cold, and "Big Bang" in contrast to various steady-state cosmologies in which nothing ever begins and nothing ever quite ends. It may seem that this archeological scenario leaves unanswered the question of how the show started and merely describes the consequences of some Great Cause that it cannot specify and does not comprehend. But really the question of how the show started answers itself: before the Big Bang there was nothing. Darkness was upon the face of the deep.

Blow-Up

NOTWITHSTANDING THE investment made by the scientific community and the general public in contemporary cosmology, a suspicion lingers that matters do not sum up as they should. Cosmologists write as if they are quite certain of the Big Bang, yet, within the last decade, they have found it necessary to augment the standard view by means of various new theories. These schemes are meant to solve problems that cosmologists were never at pains to acknowledge, so that today they are somewhat in the position of a physician reporting both that his patient has not been ill and that he has been successfully revived. The details are instructive. It is often said, for example, that the physicists Arno Penzias and Robert Wilson observed the remnants of the Big Bang when in 1962 they detected, by means of a hum in their equipment, a signal in the night sky they could not explain. This is not quite correct. What Penzias and Wilson observed was simply the same irritating and ineradicable noise that has been a feature of every electrical appliance I have ever owned. What theoreticians like Robert Dicke inferred was something else: a connection between that cackle and the cosmic background radiation released into the universe after the era of recombination.

The cosmic hum is real enough, and so, too, is the fact that the universe is bathed in background radiation. But the era of recombination is a shimmer by the doors of theory, something known indirectly and only by means of its effects. And therein lies a puzzle. Although Big Bang cosmology does predict that the universe should be bathed in a milky film of radiation, it makes no predictions about the uniformity of its temperature. Yet, looking at the sky in every direction, cosmologists have discovered that the CBR has the same temperature, to an accuracy of one part in 100,000.

Why should this be so? CBR filled the universe some 400,000 years after the Big Bang; if its temperature thereafter is utterly and entirely the same, some physical agency must have brought this about. But by the time of recombination, the Big Bang had blown up the universe to a diameter of 90,000,000 light years. A physical signal--a light beam, say- -sent hustling into the cosmos at Time Zero would, a mere 400,000 years later, be hustling still; by far the greater part of the universe would be untouched by its radiance, and so uninfluenced by the news that it carried. Since, by Einstein's theory of special relativity, nothing can travel faster than light itself, it follows that no physical agency would have had time enough to establish the homogeneity of the CBR, which appears in Big Bang cosmology as an arbitrary feature of the early universe, something that must be assumed and is not explained.

THEORIES OF inflation now make a useful ap- pearance. Their animating idea represents a contribution to cosmology from particle physics--a rare example of intellectual lend-lease, and evidence that disciplines dealing with the smallest of objects may be relevant to disciplines dealing with the largest. Within what is now known as the Standard Model, the familiar arrangement in which elementary particles are moved by various forces gives way to a mathematically more general scheme in which fields replace both particles and forces as essential theoretical structures. A field is an expanse or expression of space, something like a surface in two dimensions, or the atmosphere in three, or space and time in four. The Standard Model in particle physics consists of a dozen or so fields, exchanging energy with one another and subsidizing particles by means of the energy they contain.

It is the Higgs field that originally came to play a novel role in Big Bang cosmology. Named after Peter Higgs, the Scottish physicist who charmed it into existence, the Higgs field is purely a conjectured object. Its cosmological potential was first noted by a young American physicist, Alan Guth. Fields are structures that carry latent energy even under conditions in which the space they control is to all intents and purposes empty. The natural and stable state of the Higgs field is one in which its latent energy is at a minimum. Such is its true vacuum state, the word "vacuum" indicating that the field is empty, and the word "true" that the field is in its lowest energy configuration. But under certain physically possible circumstances, the Higgs field can find itself adventitiously trapped in a false vacuum state, a condition in which, like a spring, it is loaded with potential energy. It is thus, Guth conjectured, that the Higgs field might have found itself fluttering about the early universe, energetically throbbing and dying to be of use.

The wish is father to the act. The energy within the Higgs field is repulsive: it pushes things apart. When released, it contributes a massive jolt to the process of cosmic expansion already under way. The universe very quickly doubles in size. Space and time stretch themselves out. Particles zoom from one another. If the ordinary course of cosmic expansion is linear, inflationary expansion is exponential, like the gaunt, hollow-eyed guest gobbling the hors d'oeuvres--and everything else--at a previously decorous cocktail party. Only as the Higgs field tumbles down to its true vacuum state does inflation come to a halt, and the ordinary course of the Big Bang resume.

The mechanism of inflation, cosmologists cheerfully admit, is rather like one of those Rube Goldberg contraptions in which a door is made to open by means of a sequence that includes a flashing neon light, an insulting message in a bottle, a prizefighter wearing patent-leather shoes, and a boa constrictor with an aversion to milk. Nonetheless, they add, inflation provides a natural and plausible explanation for the fact that the CBR is uniform in temperature. If the universe under standard Big Bang cosmology is too large to allow a coordinating physical signal to reach every part of the CBR, then one redemptive idea is to cast around for a universe smaller than the standard one. This, inflation provides. Within an inflationary universe, the CBR owes its uniform temperature to the fact that it has been thoroughly mixed. At the end of the era of recombination, the CBR then surges through space like pre-warmed soup.

Inflation is an idea that has gripped the community of cosmologists. Whether it has advanced their scientific agenda is another question. As we have seen, standard Big Bang cosmology requires that features of the early universe such as the temperature of the CBR be set arbitrarily. This has seemed intellectually repugnant to many physicists: the goal of science is to reduce the arbitrariness of description. But inflationary cosmology has arbitrary features of its own, which, displaced from one corner of the theory, have a habit of popping up in another. "The need for fine-tuning of the universe," the physicist David Lindley observed of Guth's proposal, "has been obviated by fine-tuning the Higgs mechanism instead." So it has.

Soon after its introduction, Guth's model of inflation required adjustment. The Higgs field has been replaced by scalar fields, which, as Guth admits, "in many cases serve no function other than the driving of inflation." These fields must be carefully chosen if they are to do their work, a fact that Guth again honestly acknowledges: "Their nature cannot be deduced from known physics, and their detailed properties have to be hypothesized." In an interesting example of inflationary theory self-applied, inflationary fields have undergone an exponential increase of their own. Beyond mere inflation, the sort of thing that with great heartiness simply blows the universe up, there is chaotic inflation and even "eternal inflation," both of them the creations of the cosmologist Andrei Linde. Almost all cosmologists have a favored scheme; when not advancing their own, they occupy themselves enumerating the deficiencies of the others.

Red Stars at Night

STREAMING IN from space, light reaches the earth like a river rich in information, the stars in the sky having inscribed strange and secret messages on its undulations. The universe is very large, light has always whispered; the nearest galaxy to our own -- Andromeda--is more than two million light years away. But the universe has also seemed relatively static, and this, too, light suggests, the stars appearing where they have always appeared, the familiar dogs and bears and girdled archers of the constellations making their appointed rounds in the sky each night.

More than anything else, it is this impression that Big Bang cosmology rejects. The cool gray universe, current dogma holds, is a place of extraordinary violence, the galaxies receding from one another, the skin of creation stretching at every spot in space, the whole colossal structure blasting apart with terrible force. And this message is inscribed in light as well.

In one of its incarnations, light represents an undulation of the electromagnetic field; its source is the excitable atom itself, with electrons bouncing from one orbit to another and releasing energy as a result. Each atom has a spectral signature, a distinctive electromagnetic frequency. The light that streams in from space thus reveals something about the composition of the galaxies from which it was sent.

In the 1920's, the characteristic signature of hydrogen was detected in various far-flung galaxies. And then an odd discovery was made. Examining a very small sample of twenty or so galaxies, the American astronomer V. M. Slipher observed that the frequency of the hydrogen they sent into space was shifted to the red portion of the spectrum. It was an extraordinary observation, achieved by means of primitive equipment. Using a far more sophisticated telescope, Edwin Hubble made the same discovery in the late 1920's after Slipher had (foolishly) turned his attention elsewhere.

The galactic redshift, Hubble realized, was an exceptionally vivid cosmic clue, a bit of evidence from far away and long ago, and like all clues its value lay in the questions it prompted. Why should galactic light be shifted to the red and not the blue portions of the spectrum? Why, for that matter, should it be shifted at all?

An invigorating stab in the dark now followed. The pitch of a siren is altered as a police car disappears down the street, the sound waves carrying the noise stretched by the speed of the car itself. This is the familiar Doppler effect. Something similar, Hubble conjectured, might explain the redshift of the galaxies, with the distortions in their spectral signature arising as a reflection of their recessional velocity as they disappeared into the depths.

Observations and inferences resolved themselves into a quantitative relationship. The redshift of a galaxy, cosmologists affirm, and so its recessional velocity, is proportional to its distance and inversely proportional to its apparent brightness or flux. The relationship is known as Hubble's law, even though Hubble himself regarded the facts at his disposal with skepticism.

Hubble's law anchors Big Bang cosmology to the real world. Many astronomers have persuaded themselves that the law represents an observation, almost as if, peering through his telescope, Hubble had noticed the galaxies zooming off into the far distance. This is nonsense. Hubble's law consolidates a number of very plausible intellectual steps. The light streaming in from space is relieved of its secrets by means of ordinary and familiar facts, but even after the facts are admitted into evidence, the relationship among the redshift of the galaxies, their recessional velocity, and their distance represents a complicated inference, an intellectual leap.

THE BIG BANG rests on the hypothesis that the universe is expanding, and in the end the plausibility of its claims will depend on whether the universe is expanding. Astronomers can indeed point to places in the sky where the redshift of the galaxies appears to be a linear function of their distance. But in astrophysics, as in evolutionary biology, it is failure rather than success that is of significance. The astrophysical literature contains interesting and disturbing evidence that the linear relationship at the heart of Hubble's law by no means describes the facts fully.

At the end of World War II, astronomers discovered places in the sky where charged particles moving in a magnetic field sent out strong signals in the radio portion of the spectrum. Twenty years later, Alan Sandage and Thomas Mathews identified the source of such signals with optically discernible points in space. These are the quasars--quasi stellar radio sources.

Quasars have played a singular role in astrophysics. In the mid-1960's, Maarten Schmidt discovered that their spectral lines were shifted massively to the red. If Hubble's law were correct, quasars should be impossibly far away, hurtling themselves into oblivion at the far edge of space and time. But for more than a decade, the American astronomer Halton Arp has drawn the attention of the astronomical community to places in the sky where the expected relationship between redshift and distance simply fails. Embarrassingly enough, many quasars seem bound to nearby galaxies. The results are in plain sight: there on the photographic plate is the smudged record of a galaxy, and there next to it is a quasar, the points of light lined up and looking for all the world as if they were equally luminous.

These observations do not comport with standard Big Bang cosmology. If quasars have very large redshifts, they must (according to Hubble's law) be very far away; if they seem nearby, then either they must be fantastically luminous or their redshift has not been derived from their velocity. The tight tidy series of inferences that has gone into Big Bang cosmology, like leverage in commodity trading, works beautifully in reverse, physicists like speculators finding their expectations canceled by the very processes they had hoped to exploit.

Acknowledging the difficulty, some theoreticians have proposed that quasars have been caught in the process of evolution. Others have scrupled at Arp's statistics. Still others have claimed that his samples are too small, although they have claimed this for every sample presented and will no doubt continue to claim this when the samples number in the billions. But whatever the excuses, a great many cosmologists recognize that quasars mark a point where the otherwise silky surface of cosmological evidence encounters a snag.

WITHIN ANY scientific discipline, bad news must come in battalions before it is taken seriously. Cosmologists can point to any number of cases in which disconcerting evidence has resolved itself in their favor; a decision to regard the quasars with a watchful indifference is not necessarily irrational. The galaxies are another matter. They are central to Hubble's law; it is within the context of galactic observation that the crucial observational evidence for the Big Bang must be found or forged.

The battalions now begin to fill. The American mathematician I.E. Segal and his associates have studied the evidence for galactic recessional velocity over the course of twenty years, with results that are sharply at odds with predictions of Big Bang cosmology. Segal is a distinguished, indeed a great mathematician, one of the creators of modern function theory and a member of the National Academy of Sciences. He has incurred the indignation of the astrophysical community by suggesting broadly that their standards of statistical rigor would shame a sociologist. Big Bang cosmology, he writes,

owes its acceptance as a physical principle primarily to the uncritical and premature representation of [the redshift-distance relationship] as an empirical fact. . . . Observed discrepancies . . . have been resolved by a pyramid of exculpatory assumptions, which are inherently incapable of noncircular substantiation.

These are strong words of remonstration, but they are not implausible. Having constructed an elaborate scientific orthodoxy, cosmologists have acquired a vested interest in its defense. The astrophysicists J.G. Hoessell, J.E. Gunn, and T.X. Thuan, for example, report with satisfaction that within the structures described by G.O. Abell's Catalog of Bright Cluster Galaxies (1958), prediction and observation cohere perfectly to support Hubble's law. Abell's catalog is a standard astronomical resource, used by cosmologists everywhere--but it is useless as evidence for Hubble's law. "In determining whether a cluster meets selection criterion," Abell affirms, "it was assumed that their redshifts were proportional to their distance." If this is what Abell assumed, there is little point in asking what conclusions he derived. The fact that the evidence in favor of Hubble's law may be biased does not mean that it is untrue; bias may suggest nothing more than a methodological flaw. But Segal is persuaded that when the evidence is soberly considered, it does contravene accepted doctrine, statistical sloppiness functioning, as it so often does, simply to conceal the facts.

A statistical inference is compelling only if the samples upon which it rests are objectively compelling. Objectivity, in turn, requires that the process of sampling be both reasonably complete and unbiased. Segal and his colleagues have taken pains to study samples that within the limits of observation are both. Their most recent study contains a detailed parallel analysis of Hubble's law across four wave bands, one that essentially surveys all stellar objects within each band. The analysis is based on new data drawn from the G. de Vaucoleurs survey of bright cluster galaxies, which includes more than 10,000 galaxies. Hubble's own analysis, it is worthwhile to recall, was limited to twenty galaxies.

The results of their analysis are disturbing. The linear relationship that Hubble saw, Segal and his collaborators cannot see and have not found. Rather, the relationship between redshift and flux or apparent brightness that they have studied in a large number of complete samples satisfies a quadratic law, the redshift varying as the square of apparent brightness. "By normal standards of scientific due process," Segal writes, "the results of [Big Bang] cosmology are illusory."

Cosmologists have dismissed Segal's claims with a great snort of indignation. But the discrepancy from Big Bang cosmology that they reveal is hardly trivial. Like evolutionary biologists, cosmologists are often persuaded that they are in command of a structure intellectually powerful enough to accommodate gross discrepancies in the data. This is a dangerous and deluded attitude. Hubble's law embodies a general hypothesis of Big Bang cosmology--namely, that the universe is expanding--and while the law cannot be established by observation, observation can establish that it may be false. A statistically responsible body of contravening evidence has revealed something more than an incidental defect. Indifference to its implications amounts to a decision to place Big Bang cosmology beyond rational inquiry.

Monlam Chemno

SHATEVER THE facts may be, the Big Bang is also an event informed by the majesty of a great physical theory. Einstein published the equations for general relativity in 1915, and more than 80 years later, general relativity remains the only theoretical instrument remotely adequate to the representation of the universe as a whole.

General relativity is first and fundamentally an account of gravity, the force that pulls ballerinas to the ground and that fixes the planets in elliptical orbits around the sun. At the beginning of the scientific era, Isaac Newton described a universe in which space and time are absolute. The measured beating of a great clock is heard, and it is heard everywhere at once. Particles move within the unchanging vault of space. Material objects attract one another with a force proportional to their mass and inversely proportional to the square of the distance between them.

A metaphysical reorganization is required before Newton's caterpillar can emerge as Einstein's butterfly. The elements of general relativity are physical processes, a word signifying something that starts at one time and at one place and that ends at another time and another place, and so crawls along a continuum whose intrinsic structure has four dimensions. Within the arena of these physical processes, the solid structures of the Newtonian universe undergo a dissolution. The great vault of space and the uniformly beating heart dwindle and then disappear: this universe is one in which space and time have fused themselves into a single entity, and its heart is an ever-changing but reciprocating relationship between space-time and matter. Material objects direct the space and time that surround them to curve, much as a bowling ball deforms the mattress on which it rests; the curvature of space-time determines the path undertaken by physical processes, much as an ant crawling on that mattress must travel a curved path to get where it is going.

Newtonian gravity acts at a distance and as a force, the very bowels of the earth reaching to enfold an object and pull it down. But while Newton was able quantitatively to describe how gravity acts, he was unable to say why it acts at all, the aching attraction of matter for matter having no other explanation than the fact that it is so. General relativity provides an explanation of gravity in terms of the curvature of space and time. No forces are involved, and none is invoked, but gravity nonetheless emerges in this universe as a natural expression of the way the cosmos is constructed.

Freely moving objects, Einstein assumed, follow a path covering the shortest distance between points in space and time. Within the ambit of a large material object, the shortest distance between such points is curved. Ballerinas accelerate toward the center of the earth after being thrown upward by their partners because acceleration is required by the geometry in which they are embedded. In this fashion, gravity disappears as a force but remains as a fact.

If the analysis of gravity is at the center of general relativity, the intellectual tools responsible for its analysis--the equations that describe the ever-changing relationship between curvature and material objects--are responsible as well for its local character. For many years, the most precise and most interesting tests of the theory were conducted within the narrow confines of our own solar system. Cosmology, however, is a global study, one in which the universe itself is the object of contemplation, and not any of its parts. The conveyance from the local structure of the universe (the sun, the solar system) to the universe as a whole must be negotiated by a daring series of inferences.

In describing matter on a cosmic scale, cosmologists strip the stars and planets, the great galaxies and the bright bursting supernovae, of their uniqueness as places and things and replace them with an imaginary distribution: the matter of the universe is depicted as a great but uniform and homogeneous cloud covering the cosmos equitably in all its secret places. Cosmologists make this assumption because they must. There is no way to deal with the universe object by object; the equations would be inscrutable, impossible to solve. But however useful the assumption of homogeneity may be mathematically, it is false in the straightforward sense that the distribution of matter in the universe is not homogeneous at all.

Having simplified the contents of the universe, the cosmologist must take care as well, and for the same reason, to strip from the matter that remains any suggestion of particularity or preference in place. The universe, he must assume, is isotropic. It has no center whatsoever, no place toward which things tend, and no special direction or axis of coordination. The thing looks much the same wherever it is observed.

The twin assumptions that the universe is homogeneous and isotropic are not ancillary but indispensable to the hypothesis of an expanding universe; without them, no conclusion can mathematically be forthcoming. Together, these two assumptions are like the figured bass needed to chant what in Tibetan is known as Monlam Chemno, the great prayer to the cosmos.

Specification in the Dark

AN EQUATION draws the noose of an identity between two or more items. The field equations of general relativity draw that noose between curvature (the metric structure of the universe) and matter (its stress-energy tensor). But it is one thing to specify an equation, and quite another to solve it.

The mathematician in him having taken command, Einstein endeavored in 1917 to provide cosmological solutions for the field equations of his own theory. He struggled with increasing vexation. The models he was able to derive indicated that the radius of the universe was either expanding or contracting; it was a conclusion that offended his aesthetic sensibilities. By adding a parameter to his equations--the so-called cosmological constant--he was in the end able to discover a static solution, one that revealed a universe finite in extent, but unbounded, like the surface of a sphere. This static solution has a habit of dropping from the view of cosmologists, who routinely aver that Einstein's theory of general relativity uniquely specifies an expanding universe. Not so.

Einstein had hoped that the equations of general relativity would determine a single world model, or cosmic blueprint. In this he was destined to be disappointed. Months after he discovered one solution of the field equations, Willem de Sitter discovered another. In de Sitter's universe, there is no matter whatsoever, the place looking rather like a dance hall in which the music can be heard but no dancers seen, radiation filling the empty spaces and ricocheting from one end of creation to another.

In the 1920's, both Aleksandr Friedmann and Georges Lemaître discovered the solutions to the field equations that have dominated cosmology ever since, their work coming to amalgamate itself into a single denomination as Friedmann-Lemaître (FL) cosmology. Gone from their models is the cosmological constant (although it is resurrected in various inflation scenarios), and as a result the universe breathes voluptuously, its radius expanding or contracting with time.

FL cosmology does not assign to the universe a unique geometrical identity, or specify its fate forever; general relativity is mathematically compatible with a number of different physical scenarios. Like the surface of a sphere, the universe may well be closed, the whole thing falling back on itself at the end of time. A certain symmetry prevails, the life of things and all the drama of creation caught between two singularities as the universe traces a trajectory in which its initial effervescent explosion is followed by a subsequent enervating contraction.

Or yet again, the universe may well be open, space and time forever gushing into the void but with ever-decreasing intensity, like an athlete panting in shallow breaths. Such a universe is purely a mathematician's world, one seen only by the exercise of certain obscure mental muscles, and regarded by physicists (and everyone else) with glum distaste.

Or, finally, the universe may be one that occupies the Euclidean space of high-school textbooks and intuition alike, balanced precariously but balanced forever on a knife's edge between expansion and contraction.

If its predictive capacities seem unstable, FL cosmology has other peculiarities as well. Whatever the specific form its solutions take, they are alike in assigning dynamic properties to the universe as a whole. The classical distinction between the eternal vault of space and time and its entirely perishable contents has disappeared. The universe in FL cosmology is itself bound to the wheel of being, with space and time no more permanent than water and air. Light blazes, the show commences, and like some magnificent but mysterious organism, the universe expands with an exuberant rush of energy and floods nothingness with the seeds of being. In this fundamental respect, FL cosmology breaks both with tradition and with common sense.

At Time Zero

THE INTERPRETATION of a physical theory partakes of a dark art, one in which mathematical concepts are ceded dominion over the physical world. In practicing this art, the mathematician, like the necromancer that he is, is always liable to the temptation of confusing the structures over which he presides with things in the real world.

On the assumption--on the assumption--that the universe is expanding, it is irresistibly tempting to run time backward until the far-flung debris of the cosmos collects itself back into a smaller and smaller area. It seems evident, though, that this process of contraction and collapse may be continued only so far. An apple may be divided in halves, and then thirds, and then quarters, but even though the mathematical sequence of one half, one third, one fourth, and so on contains infinitely many terms, and converges ultimately to zero, the apple itself may be divided only finitely many times. This straightforward point has been the source of grave confusion. "The universe," the astronomer Joseph Silk writes, "began at time zero in a state of infinite density." It is there that (by definition) a singularity may be found. "Of course," Silk adds, "the phrase `a state of infinite density' is completely unacceptable as a physical description of the universe. . . . An infinitely dense universe [is] where the laws of physics, and even space and time, break down."

These are not words that inspire confidence. Does the phrase "a state of infinite density" describe a physical state of affairs or not? If it does, the description is uninformative by virtue of being "completely unacceptable." If it does not, the description is uninformative by virtue of being completely irrelevant. But if the description is either unacceptable or irrelevant, what reason is there to believe that the universe began in an initial singularity? Absent an initial singularity, what reason is there to believe that the universe began at all?

When prominent cosmologists tie them-selves in knots, charity tends to assign the blame to the medium in which they are navigating--books for a general audience--rather than the message they are conveying. But when it comes to the singularities, the knots form in every medium, evidence that the message is at fault and not the other way around. Cosmologists often claim that the mathematicians among them have demonstrated what they themselves may be unable clearly to express. In a passage that is typical, the astrophysicist Kip Thorne writes that "[Stephen] Hawking and [Roger] Penrose in 1970 proved--without any idealizing assumptions--that our universe must have had a space- time singularity at the beginning of its Big Bang expansion." But while it is true that Hawking and Penrose proved something, what they demonstrated remains within the gerbil wheel of mathematics; any additional inference requires a connection that the mathematician is not in a position to provide.

The concept of a singularity belongs essentially to mathematics. Singularities are not experimentally accessible objects. They cannot be weighed, measured, assessed, replicated, balanced, or seen by any modality of the senses. Within certain mathematical contexts, the concept has real content. An ordinary curve goes up one side of the blackboard and down the other; it changes its direction at a singular point. There are singularities within the calculus, and singularities in complex function theory where imaginary numbers loiter, and singularities in the space of smooth maps. There are singularities within general relativity as well, but the term covers a variety of cases, and the singularities within general relativity are distinctly odd.

In most mathematical theories a natural distinction is drawn between a figure and its background: a curve arcs within the broader ambit of an enveloping space, a mapping is easily distinguished from the spaces it connects. Typically, it is the figure that admits of a singularity: the curve changes its direction or the mapping breaks down, while the background stays the same. But in general relativity, it is the background that suffers a singularity, the very fabric of space and time giving way with a rip as curvature zooms off to infinity and space and time contort themselves. For the purposes of describing such singularities, the usual mathematical techniques are unavailing. That having been said, here is what Hawking and Penrose brought under the control of a mathematical demonstration. The setting is FL cosmology and only FL cosmology. There are three kinds of universe to consider, and innumerably many species within each type. Those that are open and forever gushing into the void are called hyperbolic. Within almost all of those hyperbolic universes, almost all processes begin at a point in the past. Within the two types of universe that remain, there is bound to be at least one process that has begun somewhere in the past.

Despite the tics--"almost all," "at least one"--the Penrose-Hawking theorems do indeed demonstrate that some universes begin in an initial singularity. But the light thrown by the Penrose-Hawking theorems flickers over a mathematical theory and so a mathematical universe. The universe that we inhabit is a physical system. Nothing but grief can come of confusing the one for the other. FL cosmology requires the existence of space-time singularities, but there is nothing in the Penrose-Hawking theorems to suggest that a space-time singularity corresponds to an explosion, or marks the beginning of an expansion, or describes an accessible portion of space and time, or connects itself to any physical state of affairs whatsoever.

Mathematical concepts achieve physical significance only when the theories in which they are embedded are confirmed by experience. If a space-time singularity is not a physical event, no such confirmation can logically be forthcoming. With the argument rolled backward, it follows that if these mathematical theories are not confirmed by experience, then neither have they achieved any physical significance.

It is Einstein who expressed the most reasonable and deeply thought views on this matter. "One may not therefore assume the validity of the [field] equation for very high density of the field and of matter," he remarked, "and one may not conclude that the beginning of the expansion must mean a singularity in the mathematical sense. All we have to realize is that the equations may not be continued over such regions."

The sharp, clean, bracing light that the Big Bang was to have thrown on the very origins of space and time lapses when it is most needed. The relevant equations of general relativity fall silent at precisely the moment we most wish they would speak.

The Closing Circle

LIKE SO many haunting human stories, the scientific story of the Big Bang is circular in the progression of its ideas and circular thus in its deepest nature. Cosmologists have routinely assumed that the universe is expanding because they have been persuaded of FL cosmology; and they have been persuaded of FL cosmology because they have routinely assumed that the universe is expanding. The pattern would be intellectually convenient if it were intellectually compelling.

If the evidence in favor of Big Bang cosmology is more suspect than generally imagined, its defects are far stronger than generally credited. Whatever else it may be, the universe is a bright, noisy, energetic place. There are monstrously large galaxies in the skies and countlessly many suns burning with fierce thermonuclear fires. Black holes are said to loiter here and there, sucking in matter and light and releasing it slowly in the form of radiation. Whence the energy for the show, the place where accounts are settled? The principles of 19th-century physics require that, in one way or another, accounts must be settled. Energy is neither created nor destroyed.

Hot Big Bang cosmology appears to be in violation of the first law of thermodynamics. The global energy needed to run the universe has come from nowhere, and to nowhere it apparently goes as the universe loses energy by cooling itself.

This contravention of thermodynamics express- es, in physical form, a general philosophical anxiety. Having brought space and time into existence, along with everything else, the Big Bang itself remains outside any causal scheme. The creation of the universe remains unexplained by any force, field, power, potency, influence, or instrumentality known to physics--or to man. The whole vast imposing structure organizes itself from absolutely nothing.

This is not simply difficult to grasp. It is incomprehensible.

Physicists, no less than anyone else, are uneasy with the idea that the universe simply popped into existence, with space and time "suddenly switching themselves on." The image of a light switch comes from Paul Davies, who uses it to express a miracle without quite recognizing that it embodies a contradiction. A universe that has suddenly switched itself on has accomplished something within time; and yet the Big Bang is supposed to have brought space and time into existence.

Having entered a dark logical defile, physicists often find it difficult to withdraw. Thus, Alan Guth writes in pleased astonishment that the universe really did arise from "essentially . . . nothing at all": as it happens, a false vacuum patch "10-26 centimeters in diameter" and "10-32 solar masses." It would appear, then, that "essentially nothing" has both spatial extension and mass. While these facts may strike Guth as inconspicuous, others may suspect that nothingness, like death, is not a matter that admits of degrees.

The attempt to discover some primordial stuff that can be described both as nothing and as something recalls the Maori contemplating the manifold mysteries of po. This apparently gives Stephen Hawking pause. "To ask what happened before the universe began," he has written, "is like asking for a point on the Earth at 91 degrees north latitude." We are on the inside of the great sphere of space and time, and while we can see to the boundaries, there is nothing beyond to see if only because there is nothing beyond. "Instead of talking about the universe being created, and maybe coming to an end," Hawking writes, "one should just say: the universe is."

Now this is a conclusion to which mystics have always given their assent; but having concluded that the universe just "is," cosmologists, one might think, would wish to know why it is. The question that Hawking wishes to evade disappears as a question in physics only to reappear as a question in philosophy; we find ourselves traveling in all the old familiar circles.

Contract

STANDING AT the gate of modern time, Isaac Newton forged the curious social pact by which rational men and women have lived ever since. The description of the physical world would be vouchsafed to a particular institution, that of mathematical physics; and it was to the physicists and not the priests, soothsayers, poets, politicians, novelists, generals, mystics, artists, astrologers, warlocks, wizards, or enchanters that society would look for judgments about the nature of the physical world. If knowledge is power, the physicists have, by this arrangement, been given an enormous privilege. But a social arrangement is among other things a contract: something is given, but something is expected as well. In exchange for their privilege, the physicists were to provide an account of the physical world at once penetrating, general, persuasive, and true.

Until recently, the great physicists have been scrupulous about honoring the terms of their contract. They have attempted with dignity to respect the distinction between what is known and what is not. Even quantum electrodynamics, the most successful theory ever devised, was described honestly by its founder, Richard Feynman, as resting on a number of unwholesome mathematical tricks.

This scrupulousness has lately been compromised. The result has been the calculated or careless erasure of the line separating disciplined physical inquiry from speculative metaphysics. Contemporary cosmologists feel free to say anything that pops into their heads. Unhappy examples are everywhere: absurd schemes to model time on the basis of the complex numbers, as in Stephen Hawking's A Brief History of Time; bizarre and ugly contraptions for cosmic inflation; universes multiplying beyond the reach of observation; white holes, black holes, worm holes, and naked singularities; theories of every stripe and variety, all of them uncorrected by any criticism beyond the trivial.

The physicists carry on endlessly because they can. Just recently, for example, Lee Smolin, a cosmologist at the University of Pennsylvania, has offered a Darwinian interpretation of cosmology, a theory of "cosmological natural selection." On Smolin's view, the Big Bang happened within a black hole; new universes are bubbling up all the time, each emerging from its own black hole and each provided with its own set of physical laws, so that the very concept of a law of nature is shown to be a part of the mutability of things.

There is, needless to say, no evidence whatsoever in favor of this preposterous theory. The universes that are bubbling up are unobservable. So, too, are the universes that have bubbled up and those that will bubble up in the future. Smolin's theories cannot be confirmed by experience. Or by anything else. What law of nature could reveal that the laws of nature are contingent? Yet the fact that when Smolin's theory is self-applied it self-destructs has not prevented physicists like Alan Guth, Roger Penrose, and Martin Rees from circumspectly applauding the effort nonetheless.

A scientific crisis has historically been the excuse to which scientists have appealed for the exculpation of damaged doctrines. Smolin is no exception. "We are living," he writes, "through a period of scientific crisis." Ordinary men and women may well scruple at the idea that cosmology is in crisis because cosmologists, deep down, have run out of interesting things to say, but in his general suspicions Smolin is no doubt correct. What we are discovering is that many areas of the universe are apparently protected from our scrutiny, like sensitive files sealed from view by powerful encryption codes. However painful, the discovery should hardly be unexpected. Beyond every act of understanding, there is an abyss.

Like Darwin's theory of evolution, Big Bang cosmology has undergone that curious social process in which a scientific theory is promoted to a secular myth. The two theories serve as points of certainty in an intellectual culture that is otherwise disposed to give the benefit of the doubt to doubt itself. It is within the mirror of these myths that we have come to see ourselves. But if the promotion of theory into myth satisfies one human agenda, it violates another. Myths are quite typically false, and science is concerned with truth. Human beings, it would seem, may make scientific theories or they may make myths, but with respect to the same aspects of experience, they cannot quite do both.

DAVID BERLINSKI, whose articles in COMMENTARY include "The Deniable Darwin" (June 1996) and "The Soul of Man Under Physics" (January 1996), is the author, most recently, of A Tour of the Calculus (Pantheon). On the Origins of Life David Berlinski

For those who are studying aspects of the origin of life, some understatement, because “it furnishes an ex- the question no longer seems to be whether life could ample of the artificial production of an organic, in- have originated by chemical processes involving non-bi- deed a so-called animal substance, from inorganic ological components but, rather, what pathway might materials.” have been followed. Wöhler’s work initiated a revolution in chem- —National Academy of Sciences (1996) istry; but it also initiated a revolution in thought. To the extent that living systems are chemical in their nature, it became possible to imagine that t is 1828, a year that encompassed the death of they might be chemical in their origin; and if I Shaka, the Zulu king, the passage in the United chemical in their origin, then plainly physical in States of the Tariff of Abominations, and the bat- their nature, and hence a part of the universe that tle of Las Piedras in South America. It is, as well, can be explained in terms of “the model for what the year in which the German chemist Friedrich science should be.”* Wöhler announced the synthesis of urea from In a letter written to his friend, Sir Joseph cyanic acid and ammonia. Hooker, several decades after Wöhler’s announce- Discovered by H.M. Roulle in 1773, urea is the chief constituent of urine. Until 1828, chemists ment, Charles Darwin allowed himself to specu- had assumed that urea could be produced only by late. Invoking “a warm little pond” bubbling up in a living organism. Wöhler provided the most con- the dim inaccessible past, Darwin imagined that vincing refutation imaginable of this thesis. His given “ammonia and phosphoric salts, light, heat, synthesis of urea was noteworthy, he observed with electricity, etc. present,” the spontaneous genera- David Berlinski, the author most recently of Infinite As- * I used this phrase, borrowed from the mathematicians J.H. Hub- cent: A Short History of Mathematics (Modern Library), bard and B.H. West, in “On the Origins of the Mind” (Commen- tary, November 2004). The idea that science must conform to a is a senior fellow of the Discovery Institute. The present essay, certain model of inquiry is familiar. Hubbard and West identify the second in a three-part series, follows “On the Origins of that model with differential equations, the canonical instruments the Mind” (November 2004), now anthologized in The Best throughout physics and chemistry. American Science Writing 2005, edited by Alan Lightman But the essentials of the model, it seems to me, lie less with the (Harper Perennial). A final installment, on the origins of particular means in which it is expressed and more with the con- matter, is scheduled for later this year. Mr. Berlinski would straints that it must meet. The idea behind the “model for what sci- ence should be” is that whatever may be a system’s initial condi- like to thank Ed Peltzer, Gustaf Arrhenius, Jonathan Wells, tions, or starting point, the laws of its development must be both Arthur Cody, Tyler Hampton, and Morris Salkoff for their unique and stable. When they are, the system that results is well helpful comments on early drafts of this essay. posed, and so a proper object of contemplation.

[22] On the Origins of Life tion of a “protein compound” might follow, with electrifying—all the more so because of the exper- this compound “ready to undergo still more com- iment’s methodological genius. Miller and Urey plex changes” and so begin Darwinian evolution it- had done nothing. Nature had done everything. self. The experiment alone had parted the cloud of un- Time must now be allowed to pass. Shall we say knowing. 60 years or so? Working independently, J.B.S. Haldane in England and A.I. Oparin in the Soviet The Double Helix Union published influential studies concerning the origin of life. Before the era of biological evo- n April 1953, just four weeks before Miller and lution, they conjectured, there must have been an I Urey would report their results in Science, James era of chemical evolution taking place in something Watson and Francis Crick published a short letter like a pre-biotic soup. A reducing atmosphere pre- in Nature entitled “A Structure for Deoxyribose vailed, dominated by methane and ammonia, in Nucleic Acid.” The letter is now famous, if only which hydrogen atoms, by donating their elec- because the exuberant Crick, at least, was persuad- trons (and so “reducing” their number), promot- ed that he and Watson had discovered the secret of ed various chemical reactions. Energy was at hand life. In this he was mistaken: the secret of life, along in the form of electrical discharges, and thereafter with its meaning, remains hidden. But in deducing complex hydrocarbons appeared on the surface of the structure of deoxyribose nucleic acid (DNA) the sea. from X-ray diffraction patterns and various chemi- The publication of Stanley Miller’s paper, “A cal details, Watson and Crick had discovered the Production of Amino Acids Under Possible Primi- way in which life at the molecular level replicates tive Earth Conditions,” in the May 1953 issue of itself. Science completed the inferential arc initiated by Formed as a double helix, DNA, Watson and Friedrich Wöhler 125 years earlier. Miller, a grad- Crick argued, consists of two twisted strings facing uate student, did his work at the instruction of each other and bound together by struts. Each Harold Urey. Because he did not contribute direct- string comprises a series of four nitrogenous bases: ly to the experiment, Urey insisted that his name adenine (A), guanine (G), thymine (T), and cyto- not be listed on the paper itself. But their work is sine (C). The bases are nitrogenous because their now universally known as the Miller-Urey experi- chemical activity is determined by the electrons of ment, providing evidence that a good deed can be the nitrogen atom, and they are bases because they its own reward. are one of two great chemical clans—the other By drawing inferences about pre-biotic evolution being the acids, with which they combine to form from ordinary chemistry, Haldane and Oparin had salts. opened an imaginary door. Miller and Urey barged Within each strand of DNA, the nitrogenous right through. Within the confines of two beakers, bases are bound to a sugar, deoxyribose. Sugar they re-created a simple pre-biotic environment. molecules are in turn linked to each other by a One beaker held water; the other, connected to the phosphate group. When nucleotides (A, G, T, or first by a closed system of glass tubes, held hydro- C) are connected in a sugar-phosphate chain, they gen cyanide, water, methane, and ammonia. The form a polynucleotide. In living DNA, two such two beakers were thus assumed to simulate the pre- chains face each other, their bases touching fingers, biotic ocean and its atmosphere. Water in the first A matched to T and C to G. The coincidence be- could pass by evaporation to the gases in the sec- tween bases is known now as Watson-Crick base ond, with vapor returning to the original alembic pairing. by means of condensation. “It has not escaped our notice,” Watson and Then Miller and Urey allowed an electrical Crick observed, “that the specific pairings we have spark to pass continually through the mixture of postulated immediately suggests a possible copying gases in the second beaker, the gods of chemistry mechanism for the genetic material” (emphasis controlling the reactions that followed with very added). Replication proceeds, that is, when a mol- little or no human help. A week after they had ecule of DNA is unzipped along its internal axis, begun their experiment, Miller and Urey discov- dividing the hydrogen bonds between the bases. ered that in addition to a tarry residue—its most Base pairing then works to prompt both strands of notable product—their potent little planet had a separated double helix to form a double helix yielded a number of the amino acids found in liv- anew. ing systems. So Watson and Crick conjectured, and so it has The effect among biologists (and the public) was proved.

[23] Commentary February 2006

The Synthesis of Protein DNA or RNA [ribonucleic acid] would think of them as templates for amino acids.” ogether with Francis Crick and Maurice Had these observations been made by anyone T Wilkins, James Watson received the Nobel but Francis Crick, they might have been regarded Prize for medicine in 1962. In his acceptance as the work of a lunatic; but in looking at any text- speech in Stockholm before the king of Sweden, book in molecular biology today, it is clear that Watson had occasion to explain his original re- Crick was simply noticing what was under his nose. search goals. The first was to account for genetic Just where are those “knobby hydrophobic sur- replication. This, he and Crick had done. The sec- faces”? To imagine that the nucleic acids form a ond was to describe the “way in which genes con- template or pattern for the amino acids is a little trol protein synthesis.” This, he was in the course like trying to imagine a glove fitting over a cen- of doing. tipede. But if the nucleic acids did not form a tem- DNA is a large, long, and stable molecule. As plate for the amino acids, then the information molecules go, it is relatively inert. It is the proteins, they contained—all of the ancient wisdom of the rather, that handle the day-to-day affairs of the cell. species, after all—could only be expressed by an in- Acting as enzymes, and so as agents of change, pro- direct form of transmission: a code of some sort. teins make possible the rapid metabolism charac- teristic of modern organisms. he idea was hardly new. The physicist Erwin Proteins are formed from the alpha-amino acids, T Schrödinger had predicted in 1945 that living of which there are twenty in living systems. The systems would contain what he called a “code prefix “alpha” designates the position of the crucial script”; and his short, elegant book, What Is Life?, carbon atom in the amino acid, indicating that it had exerted a compelling influence on every mole- lies adjacent to (and is bound up with) a carboxyl cular biologist who read it. Ten years later, the group comprising carbon, oxygen, again oxygen, ubiquitous Crick invoked the phrase “sequence hy- and hydrogen. And the proteins are polymers: like pothesis” to characterize the double idea that DNA DNA, their amino-acid constituents are formed sequences spell a message and that a code is re- into molecular chains. quired to express it. What remained obscure was But just how does the cell manage to link amino both the spelling of the message and the mecha- acids to form specific proteins? This was the prob- nism by which it was conveyed. lem to which Watson alluded as the king of Swe- The mechanism emerged first. During the late den, lost in a fog of admiration, nodded amiably. 1950’s, François Jacob and Jacques Monod ad- vanced the thesis that RNA acts as the first in a he success of Watson-Crick base pairing had chain of intermediates leading from DNA to the T persuaded a number of molecular biologists amino acids. that DNA undertook protein synthesis by the same Single- rather than double-stranded, RNA is a process—the formation of symmetrical patterns or nucleic acid: a chip from the original DNA block. “templates”—that governed its replication. After Instead of thymine (T), it contains the base uracil all, molecular replication proceeded by the divine- (U), and the sugar that it employs along its back- ly simple separation-and-recombination of match- bone features an atom of oxygen missing from de- ing (or symmetrical) molecules, with each strand of oxyribose. But RNA, Jacob and Monod argued, DNA serving as the template for another. So it was more than a mere molecule: it was a messen- seemed altogether plausible that DNA would like- ger, an instrument of conveyance, “transcribing” in wise serve a template function for the amino acids. one medium a message first expressed in another. It was Francis Crick who in 1957 first observed Among the many forms of RNA loitering in the that this was most unlikely. In a note circulated pri- modern cell, the RNA bound for duties of tran- vately, Crick wrote that “if one considers the physi- scription became known, for obvious reasons, as co-chemical nature of the amino-acid side chains, “messenger” RNA. we do not find complementary features on the nu- In transcription, molecular biologists had dis- cleic acids. Where are the knobby hydrophobic . . . covered a second fundamental process, a compan- surfaces to distinguish valine from leucine and ion in arms to replication. Almost immediately isoleucine? Where are the charged groups, in spe- thereafter, details of the code employed by the cific positions, to go with acidic and basic amino messenger appeared. In 1961, Marshall Nirenberg acids?” and J. Heinrich Matthaei announced that they had Should anyone have missed his point, Crick discovered a specific point of contact between made it again: “I don’t think that anyone looking at RNA and the amino acids. And then, in short

[24] On the Origins of Life

order, the full genetic code emerged. RNA (like they brought about a physical connection between DNA) is organized into triplets, so that adjacent se- the nucleic and the amino acids, the fact that they quences of three bases are mapped to a single were themselves nucleic acids raised a question: in amino acid. Sixty-four triplets (or codons) govern the unfolding molecular chain, just what acted to twenty amino acids. The scheme is universal, or al- adapt the adaptors to the amino acids? And this, most so. too, was a problem Crick both envisaged and The elaboration of the genetic code made possi- solved: his original suggestion mentioned both ble a remarkably elegant model of the modern cell adaptors (nucleic acids) and their enzymes (pro- as a system in which sequences of codons within teins). the nucleic acids act at a distance to determine se- And so again it proved. The act of matching quences of amino acids within the proteins: com- adaptors to amino acids is carried out by a family mands issued, responses undertaken. A third fun- of enzymes, and thus by a family of proteins: the damental biological process thus acquired molecu- aminoacyl-tRNA synthetases. There are as many lar incarnation. If replication served to divide and such enzymes as there are adaptors. The prefix then to duplicate the cell’s ancestral message, and “aminoacyl” indicates a class of chemical reactions, transcription to re-express it in messenger RNA, and it is in aminoacylation that the cargo of a car- “translation” acted to convey that message from boxyl group is bonded to a molecule of transfer messenger RNA to the amino acids. RNA. Collectively, the enzymes known as synthetases or all the boldness and power of this thesis, have the power both to recognize specific codons F the details remained on the level of what and to select their appropriate amino acid under bookkeepers call general accounting procedures. the universal genetic code. Recognition and selec- No one had established a direct—a physical—con- tion are ordinarily thought to be cognitive acts. In nection between RNA and the amino acids. psychology, they are poorly understood, but within Having noted the problem, Crick also indicated the cell they have been accounted for in chemical the shape of its solution. “I therefore proposed a terms and so in terms of “the model for what sci- theory,” he would write retrospectively, “in which ence should be.” there were twenty adaptors (one for each amino With tRNA appropriately charged, the molecule acid), together with twenty special enzymes. Each is conveyed to the ribosome, where the task of as- enzyme would join one particular amino acid to its sembling sequences of amino acids is then under- own special adaptor.” taken by still another nucleic acid, ribosomal RNA In early 1969, at roughly the same time that a (rRNA). By these means, the modern cell is at last somber Lyndon Johnson was departing the White subordinated to a rich narrative drama. To repeat: House to return to the Pedernales, the adaptors Replication duplicates the genetic message in whose existence Crick had predicted came into DNA. view. There were twenty, just as he had suggested. Transcription copies the genetic message from They were short in length; they were specific in DNA to RNA. their action; and they were nucleic acids. Collec- Translation conveys the genetic message from tively, they are now designated “transfer” RNA RNA to the amino acids—whereupon, in a fourth (tRNA). and final step, the amino acids are assembled into Folded like a cloverleaf, transfer RNA serves proteins. physically as a bridge between messenger RNA and an amino acid. One arm of the cloverleaf is called The Central Dogma the anti-coding region. The three nucleotide bases that it contains are curved around the arm’s bulb- t was once again Francis Crick, with his re- end; they are matched by Watson-Crick base pair- I markable gift for impressing his authority over ing to bases on the messenger RNA. The other end an entire discipline, who elaborated these facts into of the cloverleaf is an acceptor region. It is here what he called the central dogma of molecular bi- that an amino acid must go, with the structure of ology. The cell, Crick affirmed, is a divided king- tRNA suggesting a complicated female socket dom. Acting as the cell’s administrators, the nucle- waiting to be charged by an appropriate male ic acids embody all of the requisite wisdom—where amino acid. to go, what to do, how to manage—in the specific The adaptors whose existence Crick had pre- sequence of their nucleotide bases. Administration dicted served dramatically to confirm his hypothe- then proceeds by the transmission of information sis that such adaptors were needed. But although from the nucleic acids to the proteins.

[25] Commentary February 2006

The central dogma thus depicts an arrow mov- the origins of life that the facts engender a paradox, ing one way, from the nucleic acids to the proteins, or at least a question: in the receding molecular and never the other way around. But is anything spiral, which came first—the chicken in the form ever routinely returned, arrow-like, from its target? of DNA, or its egg in the form of various proteins? This is not a question that Crick considered, al- And if neither came first, how could life have though in one sense the answer is plainly no. Given begun? the modern genetic code, which maps four nuc- leotides onto twenty amino acids, there can be no The RNA World inverse code going in the opposite direction; an in- verse mapping is mathematically impossible. t is 1967, the year of the Six-Day war in the But there is another sense in which Crick’s cen- I Middle East, the discovery of the electroweak tral dogma does engender its own reversal. If the forces in particle physics, and the completion of a nucleic acids are the cell’s administrators, the pro- twenty-year research program devoted to the ef- teins are its chemical executives: both the staff and fects of fluoridation on dental caries in Evanston, the stuff of life. The molecular arrow goes one way Illinois. It is also the year in which Carl Woese, with respect to information, but it goes the other Leslie Orgel, and Francis Crick introduced the hy- way with respect to chemistry. pothesis that “evolution based on RNA replication Replication, transcription, and translation repre- preceded the appearance of protein synthesis” (em- sent the grand unfolding of the central dogma as it phasis added). proceeds in one direction. The chemical activities By this time, it had become abundantly clear initiated by the enzymes represent the grand un- that the structure of the modern cell was not only folding of the central dogma as it goes in the other. more complex than other physical structures but Within the cell, the two halves of the central dogma complex in poorly understood ways. And yet no combine to reveal a system of coded chemistry, an ex- matter how far back biologists traveled into the quisitely intricate but remarkably coherent tempo- tunnel of time, certain features of the modern cell ral tableau suggesting a great army in action. were still there, a message sent into the future by From these considerations a familiar figure now the last universal common ancestor. Summarizing emerges: the figure of a chicken and its egg. Repli- his own perplexity in retrospect, Crick would later cation, transcription, and translation are all under observe that “an honest man, armed with all the the control of various enzymes. But enzymes are knowledge available to us now, could only state proteins, and these particular proteins are specified that, in some sense, the origin of life appears at the by the cell’s nucleic acids. DNA requires the en- moment to be almost a miracle.” Very wisely, Crick zymes in order to undertake the work of replica- would thereupon determine never to write another tion, transcription, and translation; the enzymes re- paper on the subject—although he did affirm his quire DNA in order to initiate it. The nucleic acids commitment to the theory of “directed pansper- and the proteins are thus profoundly coordinated, mia,” according to which life originated in some each depending upon the other. Without amino- other portion of the universe and, for reasons that acyl-tRNA synthetase, there is no translation from Crick could never specify, was simply sent here. RNA; but without DNA, there is no synthesis of But that was later. In 1967, the argument pre- aminoacyl-tRNA synthetase. sented by Woesel, Orgel, and Crick was simple. If the nucleic acids and their enzymes simply Given those chickens and their eggs, something chased each other forever around the same cell, the must have come first. Two possibilities were struck result would be a vicious circle. But life has elegant- off by a process of elimination. DNA? Too stable ly resolved the circle in the form of a spiral. The and, in some odd sense, too perfect. The proteins? aminoacyl-tRNA synthetase that is required to Incapable of dividing themselves, and so, like mol- complete molecular translation enters a given cell ecular eunuchs, useful without being fecund. That from its progenitor or “maternal” cell, where it is left RNA. While it was not obviously the right specified by that cell’s DNA. The enzymes required choice for a primordial molecule, it was not obvi- to make the maternal cell’s DNA do its work enter ously the wrong choice, either. that cell from its maternal line. And so forth. The hypothesis having been advanced—if with On the level of intuition and experience, these no very great sense of intellectual confidence—bi- facts suggest nothing more mysterious than the ologists differed in its interpretation. But they did longstanding truism that life comes only from life. concur on three general principles. First: that at Omnia viva ex vivo, as Latin writers said. It is only some time in the distant past, RNA rather than when they are embedded in various theories about DNA controlled genetic replication. Second: that

[26] On the Origins of Life

Watson-Crick base pairing governed ancestral subordinated to a rational three-part sequence, be- RNA. And third: that RNA once carried on chem- ginning in the very distant past. First, the con- ical activities of the sort that are now entrusted to stituents of the cell were formed and assembled. the proteins. The paradox of the chicken and the These included the nucleotide bases, the amino egg was thus resolved by the hypothesis that the acids, and the sugars. There followed next the chicken was the egg. emergence of the ribozyme, endowed somehow The independent discovery in 1981 of the ri- with powers of self-replication. With the stage set, bozyme—a ribonucleic enzyme—by Thomas Cech a system of coded chemistry then emerged, making and Sidney Altman endowed the RNA hypothesis possible what the molecular biologist Paul Schim- with the force of a scientific conjecture. Studying mel has called “the theater of the proteins.” Thus the ciliated protozoan Tetrahymena thermophila, did matters proceed from the pre-biotic past to the Cech discovered to his astonishment a form of very threshold of the last universal common ances- RNA capable of inducing cleavage. Where an en- tor, whereupon, with inimitable gusto, life began to zyme might have been busy pulling a strand of diversify itself by means of Darwinian principles. RNA apart, there was a ribozyme doing the work This account is no longer fantasy. But it is not yet instead. That busy little molecule served not only fact. That is one reason why retracing its steps is to give instructions: apparently it took them as such an interesting exercise, to which we now turn. well, and in any case it did what biochemists had since the 1920’s assumed could only be done by an Miller Time enzyme and hence by a protein. In 1986, the biochemist Walter Gilbert was t is perhaps four billion years ago. The first of moved to assert the existence of an entire RNA I the great eras in the formation of life has com- “world,” an ancestral state promoted by the magic menced. The laws of chemistry are completely in of this designation to what a great many biologists control of things—what else is there? It is Miller would affirm as fact. Thus, when the molecular bi- Time, the period marking the transition from in- ologist Harry Noller discovered that protein syn- organic to organic chemistry. thesis within the contemporary ribosome is cat- According to the impression generally conveyed alyzed by ribosomal RNA (rRNA), and not by any in both the popular and the scientific literature, the of the familiar, old-fashioned enzymes, it appeared success of the original Miller-Urey experiment was “almost certain” to Leslie Orgel that “there once both absolute and unqualified. This, however, is was an RNA world” (emphasis added). something of an exaggeration. Shortly after Miller and Urey published their results, a number of ex- From Molecular Biology to the Origins of Life perienced geochemists expressed reservations. Miller and Urey had assumed that the pre-biotic t is perfectly true that every part of the modern atmosphere was one in which hydrogen atoms gave I cell carries some faint traces of the past. But up (reduced) their electrons in order to promote these molecular traces are only hints. By contrast, chemical activity. Not so, the geochemists con- to everyone who has studied it, the ribozyme has tended. The pre-biotic atmosphere was far more appeared to be an authentic relic, a solid and pal- nearly neutral than reductive, with little or no pable souvenir from the pre-biotic past. Its discov- methane and a good deal of carbon dioxide. ery prompted even Francis Crick to the admission Nothing in the intervening years has suggested that he, too, wished he had been clever enough to that these sour geochemists were far wrong. Writ- look for such relics before they became known. ing in the 1999 issue of Peptides, B.M. Rode ob- Thanks to the ribozyme, a great many scientists served blandly that “modern geochemistry assumes have become convinced that the “model for what that the secondary atmosphere of the primitive science should be” is achingly close to encompass- earth (i.e., after diffusion of hydrogen and helium ing the origins of life itself. “My expectation,” re- into space) . . . consisted mainly of carbon dioxide, marks David Liu, professor of chemistry and nitrogen, water, sulfur dioxide, and even small chemical biology at Harvard, “is that we will be amounts of oxygen.” This is not an environment able to reduce this to a very simple series of logical calculated to induce excitement. events.” Although often overstated, this optimism Until recently, the chemically unforthcoming is by no means irrational. Looking at the modern nature of the early atmosphere remained an em- cell, biologists propose to reconstruct in time the barrassing secret among evolutionary biologists, structures that are now plainly there in space. like an uncle known privately to dress in women’s Research into the origins of life has thus been underwear; if biologists were disposed in public to

[27] Commentary February 2006 acknowledge the facts, they did so by remarking tosine would fall to 0 percent if the reaction were that every family has one. This has now changed. extended.” The issue has come to seem troubling. A recent If the central chemical reaction favored by paper in Science has suggested that previous con- Robertson and Miller was self-defeating, it was also jectures about the pre-biotic atmosphere were se- contingent on circumstances that were unlikely. riously in error. A few researchers have argued Concentrated urea was needed to prompt their reac- that a reducing atmosphere is not, after all, quite tion; an outhouse whiff would not do. For this so important to pre-biotic synthesis as previously same reason, however, the pre-biotic sea, where imagined. concentrates disappear too quickly, was hardly the In all this, Miller himself has maintained a far place to begin—as anyone who has safely relieved more unyielding and honest perspective. “Either himself in a swimming pool might confirm with you have a reducing atmosphere,” he has written guilty satisfaction. Aware of this, Robertson and bluntly, “or you’re not going to have the organic Miller posited a different set of circumstances: in compounds required for life.” place of the pre-biotic soup, drying lagoons. In a fine polemical passage, their critic Shapiro stipu- f the composition of the pre-biotic atmosphere lated what would thereby be required: remains a matter of controversy, this can hard- I An isolated lagoon or other body of seawater ly be considered surprising: geochemists are at- would have to undergo extreme concentra- tempting to revisit an era that lies four billion years tion. . . . in the past. The synthesis of pre-biotic chemicals is It would further be necessary that the resid- another matter. Questions about them come under ual liquid be held in an impermeable vessel [in the discipline of laboratory experiments. order to prevent cross-reactions]. Among the questions is one concerning the ni- The concentration process would have to trogenous base cytosine (C). Not a trace of the be interrupted for some decades . . . to allow stuff has been found in any meteor. Nothing in the reaction to occur. comets, either, so far as anyone can tell. It is not At this point, the reaction would require buried in the Antarctic. Nor can it be produced by quenching (perhaps by evaporation to dryness) any of the common experiments in pre-biotic to prevent loss by deamination. chemistry. Beyond the living cell, it has not been At the end, one would have a batch of urea in found at all. solid form, containing some cytosine (and urea). When, therefore, M.P. Robertson and Stanley Miller announced in Nature in 1995 that they had Such a scenario, Shapiro remarked, “cannot be ex- specified a plausible route for the pre-biotic syn- cluded as a rare event on early earth, but it cannot thesis of cytosine from cyanoacetaldehyde and be termed plausible.” urea, the feeling of gratification was very consider- able. But it has also been short-lived. In a lengthy ike cytosine, sugar must also make an appear- and influential review published in the 1999 Pro- L ance in Miller Time, and, like cytosine, it too ceedings of the National Academy of Science, the New is difficult to synthesize under plausible pre-biotic York University chemist Robert Shapiro observed conditions. that the reaction on which Robertson and Miller In 1861, the German chemist Alexander Bul- had pinned their hopes, although active enough, terow created a sugar-like substance from a mixture ultimately went nowhere. All too quickly, the cyto- of formaldehyde and lime. Subsequently refined by sine that they had synthesized transformed itself a long line of organic chemists, Bulterow’s so- into the RNA base uracil (U) by a chemical reac- called formose reaction has been an inspiration to tion known as deamination, which is nothing more origins-of-life researchers ever since. mysterious than the process of getting rid of one The reaction is today initiated by an alkalizing molecule by sending it somewhere else. agent, such as thallium or lead hydroxide. There The difficulty, as Shapiro wrote, was that “the follows a long induction period, with a number of formation of cytosine and the subsequent deami- intermediates bubbling up. The formose reaction nation of the product to uracil occur[ed] at about is auto-catalytic in the sense that it keeps on going: the same rate.” Robertson and Miller had them- the carbohydrates that it generates serve to prime selves reported that after 120 hours, half of their the reaction in an exponentially growing feedback precious cytosine was gone—and it went faster loop until the initial stock of formaldehyde is ex- when their reactions took place in saturated urea. hausted. With the induction over, the formose re- In Shapiro’s words, “It is clear that the yield of cy- action yields a number of complex sugars.

[28] On the Origins of Life

Nonetheless, it is not sugars in general that are The historical task assigned to this era is a dou- wanted from Miller Time but a particular form of ble one: forming chains of nucleic acids from nuc- sugar, namely, ribose—and not simply ribose but leotides, and discovering among them those capa- dextro ribose. Compounds of carbon are naturally ble of reproducing themselves. Without the first, right-handed or left-handed, depending on how there is no RNA; and without the second, there is they polarize light. The ribose in living systems is no life. right-handed, hence the prefix “dextro.” But the In living systems, polymerization or chain-for- sugars exiting the formose reaction are racemic, mation proceeds by means of the cell’s invaluable that is, both left- and right-handed, and the yield enzymes. But in the grim inhospitable pre-biotic, of usable ribose is negligible. no enzymes were available. And so chemists have While nothing has as yet changed the funda- assigned their task to various inorganic catalysts. mental fact that it is very hard to get the right kind J.P. Ferris and G. Ertem, for instance, have report- of sugar from any sort of experiment, in 1990 the ed that activated nucleotides bond covalently when Swiss chemist Albert Eschenmoser was able to embedded on the surface of montmorillonite, a change substantially the way in which the sugars kind of clay. This example, combining technical appeared. Reaching with the hand of a master into complexity with general inconclusiveness, may the formose reaction itself, Eschenmoser altered stand for many others. two molecules by adding a phosphate group to In any event, polymerization having been con- them. This slight change prevented the formation cluded—by whatever means—the result was (in the of the alien sugars that cluttered the classical for- words of Gerald Joyce and Leslie Orgel) “a ran- mose reaction. The products, Eschenmoser re- dom ensemble of polynucleotide sequences”: long ported, included among other things a mixture of molecules emerging from short ones, like fronds ribose-2,4,-diphosphate. Although the mixture was on the surface of a pond. Among these fronds, na- racemic, it did contain a molecule close to the ri- ture is said to have discovered a self-replicating bose needed by living systems. With a few chemical molecule. But how? adjustments, Eschenmoser could plausibly claim, Darwinian evolution is plainly unavailing in this the pre-biotic route to the synthesis of sugar would exercise or that era, since Darwinian evolution be- lie open. gins with self-replication, and self-replication is It remained for skeptics to observe that Eschen- precisely what needs to be explained. But if Dar- moser’s ribose reactions were critically contingent winian evolution is unavailing, so, too, is chemistry. on Eschenmoser himself, and at two points: the The fronds comprise “a random ensemble of first when he attached phosphate groups to a num- polynucleotide sequences” (emphasis added); but ber of intermediates in the formose reaction, and no principle of organic chemistry suggests that the second when he removed them. aimless encounters among nucleic acids must lead What had given the original Miller-Urey experi- to a chain capable of self-replication. ment its power to excite the imagination was the If chemistry is unavailing and Darwin indis- sense that, having set the stage, Miller and Urey ex- posed, what is left as a mechanism? The evolution- ited the theater. By contrast, Eschenmoser remained ary biologist’s finest friend: sheer dumb luck. at center stage, giving directions and in general proving himself indispensable to the whole scene. as nature lucky? It depends on the payoff Events occurring in Miller Time would thus ap- W and the odds. The payoff is clear: an ances- pear to depend on the large assumption, still un- tral form of RNA capable of replication. Without proved, that the early atmosphere was reductive, that payoff, there is no life, and obviously, at some while two of the era’s chemical triumphs, cytosine point, the payoff paid off. The question is the odds. and sugar, remain for the moment beyond the For the moment, no one knows how precisely to powers of contemporary pre-biotic chemistry. compute those odds, if only because within the lab- oratory, no one has conducted an experiment lead- From Miller Time to Self-Replicating RNA ing to a self-replicating ribozyme. But the mini- mum length or “sequence” that is needed for a n the grand progression by which life arose contemporary ribozyme to undertake what the dis- I from inorganic matter, Miller Time has been tinguished geochemist Gustaf Arrhenius calls concluded. It is now 3.8 billion years ago. The “demonstrated ligase activity” is known. It is chemical precursors to life have been formed. A roughly 100 nucleotides. limpid pool of nucleotides is somewhere in exis- Whereupon, just as one might expect, things tence. A new era is about to commence. blow up very quickly. As Arrhenius notes, there are

[29] Commentary February 2006

4100 or roughly 1060 nucleotide sequences that are then, let us add two more: that without enzymes, 100 nucleotides in length. This is an unfathomably nucleotides were somehow formed into chains, and large number. It exceeds the number of atoms con- that by means we cannot duplicate in the laborato- tained in the universe, as well as the age of the uni- ry, a pre-biotic molecule discovered how to repro- verse in seconds. If the odds in favor of self-repli- duce itself. cation are 1 in 1060, no betting man would take them, no matter how attractive the payoff, and nei- From Self-Replicating RNA to Coded Chemistry ther presumably would nature. “Solace from the tyranny of nucleotide combi- new era is now in prospect, one that begins natorials,” Arrhenius remarks in discussing this A with a self-replicating form of RNA and ends very point, “is sought in the feeling that strict se- with the system of coded chemistry characteristic quence specificity may not be required through all of the modern cell. The modern cell—meaning one the domains of a functional oligmer, thus making a that divides its labors by assigning to the nucleic large number of library items eligible for participa- acids the management of information and to the tion in the construction of the ultimate functional proteins the execution of chemical activity. It is 3.6 entity.” Allow me to translate: why assume that billion years ago. self-replicating sequences are apt to be rare just be- It is with the advent of this era that distinctively cause they are long? They might have been quite conceptual problems emerge. The gods of chem- common. istry may now be seen receding into the distance. They might well have been. And yet all experi- The cell’s system of coded chemistry is determined ence is against it. Why should self-replicating RNA by two discrete combinatorial objects: the nucleic molecules have been common 3.6 billion years ago acids and the amino acids. These objects are dis- when they are impossible to discern under labora- crete because, just as there are no fractional sen- tory conditions today? No one, for that matter, has tences containing three-and-a-half words, there are ever seen a ribozyme capable of any form of cat- no fractional nucleotide sequences containing alytic action that is not very specific in its sequence three-and-a-half nucleotides, or fractional proteins and thus unlike even closely related sequences. No containing three-and-a-half amino acids. They are one has ever seen a ribozyme able to undertake combinatorial because both the nucleic acids and chemical action without a suite of enzymes in at- the amino acids are combined by the cell into larg- tendance. No one has ever seen anything like it. er structures. The odds, then, are daunting; and when consid- But if information management and its adminis- ered realistically, they are even worse than this al- tration within the modern cell are determined by a ready alarming account might suggest. The discov- discrete combinatorial system, the work of the cell ery of a single molecule with the power to initiate is part of a markedly different enterprise. The pe- replication would hardly be sufficient to establish riodic table notwithstanding, chemical reactions replication. What template would it replicate are not combinatorial, and they are not discrete. against? We need, in other words, at least two, The chemical bond, as Linus Pauling demonstrat- causing the odds of their joint discovery to increase ed in the 1930’s, is based squarely on quantum me- from 1 in 1060 to 1 in 10120. Those two sequences chanics. And to the extent that chemistry is ex- would have been needed in roughly the same place. plained in terms of physics, it is encompassed not And at the same time. And organized in such a way only by “the model for what science should be” but as to favor base pairing. And somehow held in by the system of differential equations that play so place. And buffered against competing reactions. conspicuous a role in every one of the great theo- And productive enough so that their duplicates ries of mathematical physics. would not at once vanish in the soundless sea. What serves to coordinate the cell’s two big In contemplating the discovery by chance of two shots of information management and chemical ac- RNA sequences a mere 40 nucleotides in length, tivity, and so to coordinate two fundamentally dif- Joyce and Orgel concluded that the requisite “li- ferent structures, is the universal genetic code. To brary” would require 1048 possible sequences. capture the remarkable nature of the facts in play Given the weight of RNA, they observed gloomi- here, it is useful to stress the word code. ly, the relevant sample space would exceed the mass By itself, a code is familiar enough: an arbitrary of the earth. And this is the same Leslie Orgel, it mapping or a system of linkages between two dis- will be remembered, who observed that “it was al- crete combinatorial objects. The Morse code, to most certain that there once was an RNA world.” take a familiar example, coordinates dashes and To the accumulating agenda of assumptions, dots with letters of the alphabet. To note that codes

[30] On the Origins of Life are arbitrary is to note the distinction between a translation are enzymatically driven, and without code and a purely physical connection between two those enzymes, a molecule of DNA or RNA would objects. To note that codes embody mappings is to do nothing whatsoever. Contrary to what Woese embed the concept of a code in mathematical lan- imagines, no fundamental physical principles ap- guage. To note that codes reflect a linkage of some pear directly at work anywhere in the modern cell. sort is to return the concept of a code to its human The most difficult and challenging problem as- uses. sociated with the origins of life is now in view. One In every normal circumstance, the linkage half of the modern system of coded chemistry—the comes first and represents a human achievement, genetic code and the sequences it conveys—is, something arising from a point beyond the coding from a chemical perspective, arbitrary. The other system. (The coordination of dot-dot-dot-dash- half of the system of coded chemistry—the activity dash-dash-dot-dot-dot with the distress signal of the proteins—is, from a chemical perspective, S-O-S is again a familiar example.) Just as no word necessary. In life, the two halves are coordinated. explains its own meaning, no code establishes its The problem follows: how did that—the whole sys- own nature. tem—get here? The conceptual question now follows. Can the origins of a system of coded chemistry be ex- he prevailing opinion among molecular bi- plained in a way that makes no appeal whatsoever T ologists is that questions about molecular-bi- to the kinds of facts that we otherwise invoke to ological systems can only be answered by molecu- explain codes and languages, systems of communi- lar-biological experiments. The distinguished mol- cation, the impress of ordinary words on the world ecular biologist Horoaki Suga has recently demon- of matter? strated the strengths and the limitations of the ex- In this regard, it is worth recalling that, as Hu- perimental method when confronted by difficult bert Yockey observes in Information Theory, Evolu- conceptual questions like the one I have just posed. tion, and the Origin of Life (2005), “there is no trace The goal of Suga’s experiment was to show that in physics or chemistry of the control of chemical a set of RNA catalysts (or ribozymes) could well reactions by a sequence of any sort or of a code be- have played the role now played in the modern cell tween sequences.” by the protein family of aminoacyl synthetases. Until his work, Suga reports, there had been no riting in the 2001 issue of the journal convincing demonstration that a ribozyme was able W RNA, the microbiologist Carl Woese re- to perform the double function of a synthetase— ferred ominously to the “dark side of molecular bi- that is, recognizing both a form of transfer RNA ology.” DNA replication, Woese wrote, is the ex- and an amino acid. But in Suga’s laboratory, just traordinarily elegant expression of the structural such a molecule made a now-celebrated appear- properties of a single molecule: zip down, divide, ance. With an amino acid attached to its tail, the ri- zip up. The transcription into RNA follows suit: bozyme managed to cleave itself and, like a snake, copy and conserve. In each of these two cases, affix its amino-acid cargo onto its head. What is structure leads to function. But where is the coor- more, it could conduct this exercise backward, dinating link between the chemical structure of shifting the amino acid from its head to its tail DNA and the third step, namely, translation? again. The chemical reactions involved acylation: When it comes to translation, the apparatus is precisely the reactions undertaken by synthetases baroque: it is incredibly elaborate, and it does not in the modern cell. reflect the structure of any molecule. Horoaki Suga’s experiment was both interesting These reflections prompted Woese to a somber and ingenious, prompting a reaction perhaps best conclusion: if “the nucleic acids cannot in any way expressed as, “Well, would you look at that!” It has recognize the amino acids,” then there is no “fun- altered the terms of debate by placing a number of damental physical principle” at work in translation new facts on the table. And yet, as so often hap- (emphasis added). pens in experimental pre-biotic chemistry, it is by But Woese’s diagnosis of disorder is far too par- no means clear what interpretation the facts will tial; the symptoms he regards as singular are in fact sustain. widespread. What holds for translation holds as Do Suga’s results really establish the existence of well for replication and transcription. The nucleic a primitive form of coded chemistry? Although un- acids cannot directly recognize the amino acids expected in context, the coordination he achieved (and vice versa), but they cannot directly replicate or between an amino acid and a form of transfer RNA transcribe themselves, either. Both replication and was never at issue in principle. The question is

[31] Commentary February 2006 whether what was accomplished in establishing a means of mistakes in replication, it introduces the chemical connection between these two molecules possibility of variety into the biological world. On was anything like establishing the existence of a the assumption that subsequent changes to the sys- code. If so, then organic chemistry itself could prop- tem follow a law of increasing marginal utility, one erly be described as the study of codes, thereby can then envisage the eventual emergence of a sys- erasing the meaning of a code as an arbitrary map- tem of coded chemistry—a system that can be ex- ping between discrete combinatorial objects. plained in terms of “the model for what science Suga, in summarizing the results of his research, should be.” captures rhetorically the inconclusiveness of his It was no doubt out of considerations like these achievement. “Our demonstration indicates,” he that, in coming up against what he called the “dark writes, “that catalytic precursor tRNA’s could have side of molecular biology,” Carl Woese was con- provided the foundation of the genetic coding sys- cerned to urge upon the biological community the tem.” But if the association at issue is not a code, benefits of “an all-out Darwinian perspective.” But however primitive, it could no more be the “foun- the difficulty with “an all-out Darwinian perspec- dation” of a code than a feather could be the foun- tive” is that it entails an all-out Darwinian impedi- dation of a building. And if it is the foundation of ment: notably, the assignment of a degree of fore- a code, then what has been accomplished has been sight to a Darwinian process that the process could accomplished by the wrong agent. not possibly possess. In Suga’s experiment, there was no sign that the The hypothesis of an RNA world trades bril- execution of chemical routines fell under the con- liantly on the idea that a divided modern system trol of a molecular administration, and no sign, ei- had its roots in some form of molecular symmetry ther, that the missing molecular administration had that was then broken by the contingencies of life. anything to do with executive chemical routines. At some point in the transition to the modern sys- The missing molecular administrator was, in fact, tem, an ancestral form of RNA must have assigned Suga himself, as his own account reveals. The rele- some of its catalytic properties to an emerging fam- vant features of the experiment, he writes, ily of proteins. This would have taken place at a “allow[ed] us to select active RNA molecules with given historical moment; it is not an artifact of the selectivity toward a desired amino acid” (emphasis imagination. Similarly, at some point in the transi- added). Thereafter, it was Suga and his collabora- tion to a modern system, an ancestral form of RNA tors who “applied stringent conditions” to the ex- must have acquired the ability to code for the cat- periment, undertook “selective amplification of the alytic powers it was discarding. And this, too, must self-modifying RNA molecules,” and “screened” have taken place at a particular historical moment. vigorously for “self-aminoacylation activity” (em- The question, of course, is which of the two steps phasis added throughout). came first. Without life acquiring some degree of foresight, neither step can be plausibly fixed in f nothing else, the advent of a system of coded place by means of any schedule of selective advan- I chemistry satisfied the most urgent of impera- tages. How could an ancestral form of RNA have tives: it was needed and it was found. It was needed acquired the ability to code for various amino acids because once a system of chemical reactions reach- before coding was useful? But then again, why es a certain threshold of complexity, nothing less should “ribozymes in an RNA world,” as the mole- than a system of coded chemistry can possibly mas- cular biologists Paul Schimmel and Shana O. Kel- ter the ensuing chaos. It was found because, after ley ask, “have expedited their own obsolescence?” all, we are here. Could the two steps have taken place simultane- Precisely these circumstances have persuaded ously? If so, there would appear to be very little many molecular biologists that the explanation for difference between a Darwinian explanation and the emergence of a system of coded chemistry must the frank admission that a miracle was at work. If in the end lie with Darwin’s theory of evolution. As no miracles are at work, we are returned to the one critic has observed in commenting on Suga’s place from which we started, with the chicken-and- experiments, “If a certain result can be achieved by egg pattern that is visible when life is traced back- direction in a laboratory by a Suga, surely it can ward now appearing when it is traced forward. also be achieved by chance in a vast universe.” It is thus unsurprising that writings embodying A self-replicating ribozyme meets the first con- Woese’s “all-out Darwinian perspective” are domi- dition required for Darwinian evolution to gain nated by references to a number of unspecified but purchase. It is by definition capable of replication. mysteriously potent forces and obscure condition- And it meets the second condition as well, for, by al circumstances. I quote without attribution be-

[32] On the Origins of Life cause the citations are almost generic (emphasis ology or on Darwinian adaptations. His vision added throughout): treats the living system as more fundamental than . its particular species, claiming to represent the The aminoacylation of RNA initially must “universal and deterministic features of any system have provided some selective advantage. of chemical interactions based on a water-covered . The products of this reaction must have con- but rocky planet such as ours.” ferred some selective advantage. This view of things—metabolism first, as it is often called—is not only intriguing in itself but is . However, the development of a crude mech- enhanced by a firm commitment to chemistry and anism for controlling the diversity of possible to “the model for what science should be.” It has peptides would have been advantageous. been argued with great vigor by Morowitz and oth- . [P]rogressive refinement of that mechanism ers. It represents an alternative to the RNA world. would have provided further selective advantage. It is a work in progress, and it may well be right. Nonetheless, it suffers from one outstanding de- And so forth—ending, one imagines, in reduc- fect. There is as yet no evidence that it is true. tion to the all-purpose imperative of Darwinian theory, which is simply that what was must have t is now more than 175 years since Friedrich been. I Wöhler announced the synthesis of urea. It would be the height of folly to doubt that our un- Now It Is Now derstanding of life’s origins has been immeasurably improved. But whether it has been immeasurably t the conclusion of a long essay, it is custom- improved in a way that vigorously confirms the A ary to summarize what has been learned. In daring idea that living systems are chemical in their the present case, I suspect it would be more pru- origin and so physical in their nature—that is an- dent to recall how much has been assumed: other question entirely. First, that the pre-biotic atmosphere was chemi- In “On the Origins of the Mind,” I tried to show cally reductive; second, that nature found a way to that much can be learned by studying the issue synthesize cytosine; third, that nature also found a from a computational perspective. Analogously, in way to synthesize ribose; fourth, that nature found contemplating the origins of life, much—in fact, the means to assemble nucleotides into polynu- more—can be learned by studying the issue from cleotides; fifth, that nature discovered a self-repli- the perspective of coded chemistry. In both cases, cating molecule; and sixth, that having done all however, what seems to lie beyond the reach of that, nature promoted a self-replicating molecule “the model for what science should be” is any suc- into a full system of coded chemistry. cess beyond the local. All questions about the These assumptions are not only vexing but pro- global origins of these strange and baffling systems gressively so, ending in a serious impediment to seem to demand answers that the model itself can- thought. That, indeed, may be why a number of bi- not by its nature provide. ologists have lately reported a weakening of their It goes without saying that this is a tentative commitment to the RNA world altogether, and a judgment, perhaps only a hunch. But let us suppose desire to look elsewhere for an explanation of the that questions about the origins of the mind and emergence of life on earth. “It’s part of a quiet par- the origins of life do lie beyond the grasp of “the adigm revolution going on in biology,” the bio- model for what science should be.” In that case, we physicist Harold Morowitz put it in an interview in must either content ourselves with its limitations or New Scientist, “in which the radical randomness of revise the model. If a revision also lies beyond our Darwinism is being replaced by a much more sci- powers, then we may well have to say that the mind entific law-regulated emergence of life.” and life have appeared in the universe for no very Morowitz is not a man inclined to wait for the good reason that we can discern. details to accumulate before reorganizing the vista Worse things have happened. In the end, these of modern biology. In a series of articles, he has ar- are matters that can only be resolved in the way gued for a global vision based on the biochemistry that all such questions are resolved. We must wait of living systems rather than on their molecular bi- and see.

[33] The deniable Darwin Berlinski, David Commentary; Jun 1996

A Scientific Scandal David Berlinski

IN SCIENCE, as in life, it is always an excellent idea Nilsson and Pelger's paper has, for understand- to cut the cards after the deck has been shuf- able reasons, been widely circulated and widely fled. One may admire the dealer, but trust is an- praised, and in the literature of evolutionary biolo- other matter. gy it is now regularly cited as definitive. Not the In a recent essay in COMMENTARY, "Has Darwin least of its remarkable authority is derived from the Met His Match?" (December 2002), I discussed, belief that it contains, in the words of one of its de- evaluated, and criticized theories of intelligent de- fenders, a "computer simulation of the eye's evolu- sign, which have presented the latest challenge to tion." Darwin's theory of evolution. In the course of the If this were true, it would provide an extremely discussion I observed that the evolution of the important defense of Darwin's theory. Although a mammalian eye has always seemed difficult to computer simulation is not by itself conclusive-a imagine. It is an issue that Darwin himself raised, simulation is one thing, reality another-it is often and although he settled the matter to his own satis- an important link in an inferential chain. In the faction, biologists have long wished for a direct case of Darwin's theory, the matter is especially demonstration that something like a functional eye pressing since in the nature of things the theory could be formed in reasonable periods of time by cannot be confirmed over geological time by any means of the Darwinian principles of random vari- experimental procedure, and it has proved very dif- ation and natural selection. ficult to confirm under laboratory conditions. The Just such a demonstration, I noted in my essay, claim that the eye's evolution has been successfully is what the biologists Dan-Erik Nilsson and Su- simulated by means of Darwinian principles, with sanne Pelger seemed to provide in a 1994 paper. results falling well within time scales required by Given nothing more than time and chance, a "light- the theory, is thus a matter of exceptional scientific sensitive patch," they affirmed, can "gradually turn importance. into a focused-lens eye," and in the space of only a And not just scientific importance, I might add; so few hundred thousand years-a mere moment, as dramatic a confirmation of Darwinian theory car- such things go. ries large implications for our understanding of the human species and its origins. This is no doubt DAVID BERLINSKI is the author of A Tour of the Calcu- lus, The Advent of the Algorithm, and Newton's Gift. 1 "A Pessimistic Estimate of the Time Required for an Eye to Evolve," Proceedings of the Royal Society, London B (1994) 256, 53- His new book, Secrets of the Vaulted Sky, isforthcoming 58. In my essay I twice misspelled Susanne Pelger's name, for which from Harcourtlater this year I apologize.

[29] COMMENTARY APRIL 2003 why the story of Nilsson and Pelger's computer What they do assume, if only implicitly, is that simulation has spread throughout the world. Their changes to the initial patch involve either a defor- study has been cited in essays, textbooks, and pop- mation of its shape or a thickening of its cells. The ular treatments of Darwinism like River Out of Eden patch can be stretched, dimpled, and pulled or by the famous Oxford evolutionist Richard Daw- pushed around, and cells may move closer to one an- kins; accounts of it have made their way onto the other, like bond salesmen converging on a customer. Internet in several languages; it has been promot- So much for what changes. What is the change ed to the status of a certainty and reported as fact worth? Assuming (reasonably enough) that an eye in the press, where it is inevitably used to champi- is an organ used in order to see, Nilsson and Pel- on and vindicate Darwin's theory of evolution. ger represent its value to an organism by a single In my essay, I suggested that Nilsson and Pelger's quantitative character or function, which they des- arguments are trivial and their conclusions unsub- ignate as "spatial resolution" or "visual acuity"- stantiated. I also claimed that representations of sharp sight, in short. Visual acuity confers an ad- their paper by the scientific community have in- vantage on an organism, and so, in any generation, volved a serious, indeed a flagrant, distortion of their natural selection "constantly favors an increase in work. But in a letter published in the March issue of the amount of detectable spatial information." COMMENTARY, the physicist Matt Young, whom I There are two ways in which visual acuity may singled out for criticism (and whose words I have be increased in an initial light-sensitive patch: a) by quoted here), repeated and defended his characteri- the "invagination" of the patch, so that it becomes zation of Nilsson and Pelger's work as a "computer progressively more concave and eventually forms simulation of the eye's evolution." It is therefore the enclosed interior of a sphere; and b) by the necessary to set the matter straight in some detail. constriction of the sphere's aperture (the two I hope this exercise will help to reveal, with a rounded boundaries formed as the flat patch un- certain uncomfortable clarity, just how scientific dergoes invagination). These changes may be rep- orthodoxy works, and how it imposes its opinions resented on sheets of high-school graph paper on on the faithful. which two straight lines-the x and y axes of the system-have been crossed. On the first sheet, rep- H ERE IN their own words is the main argument resenting invagination, visual acuity moves upward H 1 of Nilsson and Pelger's paper: on one axis as invagination moves to the right on the other; on the second sheet, visual acuity moves Theoretical considerations of eye design allow upward as constriction moves to the right. The us to find routes along which the optical struc- curves that result, Nilsson and Pelger assert, are ture of the eye may have evolved. If selection continuous and increasing. They do not hurdle constantly favors an increase in the amount of over any gaps, and they go steadily upward until detectable spatial information, a light-sensitive they reach a theoretical maximum. patch will gradually turn into a focused-lens The similar shape of the two graphs notwith- eye through continuous small improvements in standing, invagination and aperture constriction design. An upper limit for the number of gen- exercise different effects on visual acuity. "Initially, erations required for the complete transforma- deepening of the pit"-i.e., invagination-"is by far tion can be calculated with a minimum number the most efficient strategy," Nilsson and Pelger pes- of assumptions. Even with a consistently write; "but when the pit depth equals the width, becomes simistic approach, the time required becomes more efficient than thousand aperture constriction amazingly short: only a few hundred of the pit." From this, they years. continued deepening conclude that natural selection would act "first to And here is how they arrived at their conclusions. favor depression and invagination of the light-sen- The setting is "a single circular patch of light-sensi- sitive patch, and then gradually change to favor tive cells"-a retina, in effect--"which is bracketed constriction of the aperture." and surrounded by dark pigment." A "protective lay- er" lies above these light-sensitive cells, so that the T HE RESULT is a pin-hole eye, which is surely pigment, the light-sensitive cells, and the protective an improvement on no eye at all. But there ex- layer form a kind of sandwich. Concerning the light- ists an aperture size beyond which visual acuity can- sensitive patch itself, Nilsson and Pelger provide no not be improved without the introduction of a lens. further details, indicating neither its size nor the Having done all that it can do, the pin-hole eye laps- number of cells it might contain. es. Cells within the light-sensitive sphere now oblig-

[301 A SCIENTIFIC SCANDAL ingly begin to thicken themselves, bringing about a that their initial light-sensitive patch changes in "local increase" in the eye's refractive index and so 1-percent steps. They illustrate the procedure with forming a lens. When the focal length of the lens is the example of a flat one-foot ruler that also changes 2.55 times its radius-the so-called Mattiessen in 1-percent steps. After the first step, the ruler will ratio-the eye will have achieved, Nilsson and Pel- be one foot plus 1 percent of one foot long; after ger write, the "ideal solution for a graded-index lens the second step, it will be 1-percent longer than the with a central refractive index of 1.52."2 length just achieved; and so forth. It requires rough- Thereafter, the lens "changes its shape from el- ly 70 steps to double a one-foot ruler in length. lipsoid to spherical and moves to the center of cur- Putting the matter into symbols, 1.0170 2. vature of the retina." A flat iris "gradually forms by Nilsson and Pelger undertake a very similar cal- stretching of the original aperture," while the culation with respect to their initial light-sensitive "focal length of the lens .. . gradually shortens, patch. But since the patch is a three-dimensional [until] it equals the distance to the retina ... pro- object, they are obliged to deal with three dimen- ducing a sharply focused image." The appearance sions of change. Growing in steps of 1 percent, of this spherical, graded-index lens, when placed in their blob increases its length, its curvature, and its the center of curvature of the retina, produces "vir- volume. When all of these changes are shoe- tually aberration-free imaging over the full 180 de- horned together, the patch will have increased in grees of the visual field." magnitude along some overall (but unspecified) di- The same assumptions that governed invagina- mension. tion and aperture constriction hold sway here as The chief claim of their paper now follows: to well. Plotted against increasing lens formation, vi- achieve the visual acuity that is characteristic of a sual acuity moves smoothly and steadily upward as "focused camera-type eye with the geometry typi- a graded-index lens makes its appearance, changes cal for aquatic animals," it is necessary that an ini- its shape, and moves to center stage. When these tial patch be made 80,129,540 times larger (or transformations have been completed, the result is greater or grander) than it originally was. This a "focused camera-type eye with the geometry typ- number represents the magnitude of the blob's in- ical for aquatic animals." crease in size. How many steps does that figure rep- One step remains. Nilsson and Pelger now amal- resent? Since 80,129,540 = 1.011,829, Nilsson and gamate invagination, constriction, and lens form- Pelger conclude that "altogether 1,829 steps of 1 ation into a single "transformation," which they percent are required" to bring about the requisite represent by juxtaposing, against changes in visual transformation. acuity, changes to the original patch in increments These steps, it is important to remember, do not of 1 percent. The resulting curve, specifying quan- represent temporal intervals. We still need to assess titatively how much visual acuity may be purchased how rapidly the advantages represented by such a for each 1-percent unit of change, is ascending, in- transformation would spread in a population of or- creasing, and straight, rising like an arrow at an ganisms, and so answer the question of how long angle of roughly 45 degrees from its point of ori- the process takes. In order to do this, Nilsson and gin. Transformations are "optimal" in the sense Pelger turn to population genetics. The equation that they bring about as much visual acuity as the- that follows involves the multiplication of four oretically possible, with the "geometry of each numbers: stage [setting] an upper limit to the spatial resolu- tion of the eye." R = hxix Vxm It is the existence and shape of this fourth curve Here, R is the response (i.e. visual acuity in each that justify their claim that "a light-sensitive patch generation), h is the coefficient of heredity, i des- will gradually turn into a focused-lens eye through ignates the intensity of selection, Vis the coeffi- continuous small improvements in design" (empha- cient of variation (the ratio of the standard devia- sis added). This is not the happiest formulation tion to the mean), and m, the mean value for visual they could have chosen. acuity. These four numbers designate the extent to which heredity is responsible for visual acuity, the How MUCH does the initial light-sensitive intensity with which selection acts to prize it, the patch have to change in order to realize a way its mean or average value is spread over a pop- focused camera-type eye? And how long will it take 2 A graded-index lens is a lens that is not optically homogeneous; to do so? These are the questions now before us. the figure of 1.52 is "the value close to the upper limit for biologi- As I have mentioned, Nilsson and Pelger assume cal material."

[3 1] COMMENTARY APRIL 2003 ulation, and the mean or average value itself. Val- claimed, of certain elaborate calculations. But no ues are assigned as estimates to the first three details are given either in their paper or in its bibli- numbers; the mean is left undetermined, rising ography. The calculations to which they allude re- through each generation. main out of sight, if not out of mind. As for the estimates themselves, Nilsson and Pel- The 1-percent steps: in what units are they ex- ger assume that h2 = .50; that i = 0.01; and that V= 0.01. pressed? And how much biological change is rep- On this basis, they conclude that R=0.00005mn. resented by each step? Nilsson and Pelger do not The response in each new generation of light-sen- say. Nor do they coordinate morphological change, sitive patches is 0.00005 times the mean value of which they treat as simple, with biochemical visual acuity in the previous generation of light- change, which in the case of light sensitivity is sensitive patches. known to be monstrously complex. Their overall estimate-the conclusion of their Does invagination represent a process in which paper-now follows in two stages. Assume that n the patch changes as a whole, like a balloon being represents the number of generations required to dimpled, or is it the result of various local process- transform a light-sensitive patch into a "focused es going off independently as light-sensitive cells camera-type eye with the geometry typical for jostle with one another and change their position? aquatic animals." (In small aquatic animals, a gen- Are the original light-sensitive cells the complete eration is roughly a year.) If, as we have seen, package, or are new light-sensitive cells added to the mean value of visual acuity of such an eye is the ensemble as time proceeds? Do some cells lose 1.011829= 80,129,540, where 1,829 represents the their sensitivity and get out of the light-sensing number of steps required and 80,129,540 describes business altogether? We do not know, because the extent of the change those steps bring about; Nilsson and Pelger do not say. and if 1.00005" = 1.011829 = 80,129,540, then it fol- Biologists commenting on Darwin's theory have lows that n = 363,992. almost always assumed that evolution reflects what It is this figure-363,992-that allows Nilsson the French biologist Francois Jacob called brico- and Pelger to conclude at last that "the time re- lage-a process of tinkering. Biological structures quired [is] amazingly short: only a few hundred are put together out of pieces; they adapt their thousand years." And this also completes my expo- function to changes in their circumstances; they get sition of Nilsson and Pelger's paper. Business be- by. This suggests that in the case of eye formation, fore pleasure. morphological change might well purchases less vi- sual acuity than Nilsson and Pelger assume, the eye N ILSSON AND Pelger's work is a critic's smor- being tinkered into existence instead of flogged up gasbord. Questions are free and there are an adaptive peak. But if, say, only half as much vi- second helpings. sual acuity is purchased for each of Nilsson and Every scientific paper must begin somewhere. Pelger's 1-percent steps, twice as many steps will be Nilsson and Pelger begin with their assumption needed to achieve the effect they claim. What is that, with respect to the eye, morphological change their justification for the remarkably strong asser- comes about by invagination, aperture constriction, tion that morphological transformations purchase and lens formation. Specialists may wish to know an optimal amount of visual acuity at each step? where those light-sensitive cells came from and Again we do not know, because they do not say. why there are no other biological structures coor- More questions-and we have not even finished dinated with or contained within the interior of the the hors d'oeuvres. The plausibility of Nilsson and initial patch-for example, blood vessels, nerves, or Pelger's paper rests on a single number: 1,829. But bones. But these issues may be sensibly deferred. without knowing precisely how the number 1,829 Not so the issues that remain. Nilsson and Pel- has been derived, the reader has no way of deter- ger treat a biological organ as a physical system, mining whether it is reasonable or even meaningful. one that is subject to the laws of theoretical optics. If nothing else, the number 1,829 represents the There is nothing amiss in that. But while theoreti- maximum point of a curve juxtaposing visual acu- cal optics justifies a qualitative relationship between ity against morphological transformation. Now, a visual acuity on the one hand and invagination, respect for the ordinary mathematical decencies aperture constriction, and lens formation on the would suggest that the curve is derived from the other, the relationships that Nilsson and Pelger number, and the number from various calculations. specify are tightly quantitative. Numbers make an But all such calculations are missing from Nilsson appearance in each of their graphs: the result, it is and Pelger's paper. And if the calculations are not

[32] A SCIENTIFIC SCANDAL given, neither are any data. Have Nilsson and Pel- step could well require half again or twice the ger, for example, verified their estimate, either by number of generations they suggest. Why do Nils- showing that 1,829 1-percent steps do suffice to son and Pelger match steps to generations in the transform a patch into an eye, or by showing that way they do? I have no idea, and they do not say. such an eye may, in 1,829 -percent steps, be re- solved backward into an initial light-sensitive E ARE at last at the main course. Curiously patch? Once again, we do not know because they enough, it is the intellectual demands im- do not say. posed by Darwin's theory of evolution that serve to Still other questions suggest themselves. Al- empty Nilsson and Pelger's claims of their remain- though natural selection is mentioned by Nilsson ing plausibility. and Pelger, it is a force that plays no role in their Nilsson and Pelger assert that only 363,992 gen- reasoning. Beyond saying that it "constantly favors erations are required to generate an eye from an an increase in the amount of detectable spatial in- initial light-sensitive patch. As I have already ob- formation," they say nothing at all. This is an ig- served, the number 363,992 is derived from the num- nominious omission in a paper defending Darwin- ber 80,129,540, which is derived from the number ian principles. An improvement in visual acuity is 1,829-which in turn is derived from nothing at no doubt a fine thing for an organism; but no form all. Never mind. Let us accept 1,829 pour le sport. If of biological change is without cost. Nilsson and Pelger intend their model to be a vin- Let us agree that in the development of an eye, dication of Darwin's theory, then changes from one an initial light-sensitive patch in a given organism step to another must be governed by random becomes invaginated over time. Such a change re- changes in the model's geometry, followed by some quires a corresponding structural change to the or- mechanism standing in for natural selection. These ganism's anatomy. If nothing else, the development are, after all, the crucial features of any Darwinian of an eye requires the formation of an eye socket- theory. But in their paper there is no mention hardly a minor matter in biological terms. Is it re- whatsoever of randomly occurring changes, and nat- ally the case that an organism otherwise adapted to ural selection plays only a ceremonial role in their its environment would discover that the costs in- deliberations. volved in the reconstruction of its skull are nicely At the beginning of their paper, Nilsson and Pel- balanced by what would initially be a very modest ger write of their initial light-sensitive patch that improvement in sensitivity to light? I can imagine "we expose this structure to selection pressure fa- the argument going either way, but surely an argu- voring spatial resolution" (emphasis added), and ment is needed. later that "[a]s the lens approaches focused condi- Then there is Nilsson and Pelger's data-free tions, selection pressure gradually appears to ... ad- way with statistics. What is the basis of the mathe- just its size to agree with Mattiesen's ratio" (em- matical values chosen for the numbers they use in phasis added). But whatever Nilsson and Pelger assessing how rapidly transformation spreads in a may have been doing to their patch, they have not population of eye patches? The coefficient of vari- been exposing it to "selection pressure." The patch ation is the ratio of the standard deviation to the does only what they have told it to do. By the same mean. The standard deviation, one might ask, of token, selection pressures play no role in adjusting what? No population figures are given; there are the size of their lenses to agree with Mattiesen's no quantitative estimates of any relevant numeri- ratio. That agreement is guaranteed, since it is cal parameter. Why is selection pressure held con- Nilsson and Pelger who bring it about, drawing the stant over the course of 300,000 years or so, when curve and establishing the relevant results. What plainly the advantages to an organism of increas- Nilsson and Pelger assume is that natural selection ing light sensitivity will change at every step up the would track their results; but this assumption is adaptive slope? Why do they call their estimates never defended in their paper, nor does it play the pessimistic (that is, conservative) rather than wild- slightest role in their theory. ly optimistic? And for an obvious reason: if there are no random Finally, Nilsson and Pelger offer an estimate of variations occurring in their initial light-sensitive the number of steps, computed in -percent (actu- patch, then natural selection has nothing to do. And ally, 1.00005-percent) intervals, that are required to there are no random variations in that patch, their transform their initial patch. At one point, they model succeeding as a defense of Darwin's theory convert the steps into generations. But a step is not only by first emptying the theory of its content. a temporal unit, and, for all anyone knows, each An example may make clearer both the point

[33] COMMENTARY APRIL 2003 and its importance. Only two steps are required to understanding multi-component models and their change the English word "at" to the English word complex interrelationships." "do": "at" to "ao" and "ao" to "do." The changes Whatever the merits of computer simulation, are obvious: they have been designed to achieve the however, they are beside the point in assessing specified effect. But such design is forbidden in Nilsson and Pelger's work. In its six pages, their Darwinian theory. So let us say instead, as Darwin paper contains no mention of the words "comput- must, that letters are chosen randomly, for instance er" or "simulation." There are no footnotes indi- by being fished from an urn. In that case, it will cating that a computer simulation of their work ex- take, on average, 676 changes (26 letters times 26) ists, and their bibliography makes no reference to to bring about the same two steps. any work containing such a simulation. Similarly, depending on assessments of probabili- Curious about this point, I wrote to Dan-Erik ty, the number of changes required to bring about a Nilsson in the late summer of 2001. "Dear David," single step in Nilsson and Pelger's theory may range he wrote back courteously and at once, widely. It may, in fact, be anything at all. How long with Pelger is not would it take to transform a light-sensitive patch You are right that my article on computer simulation of eye evolu- into a fully functioning eye? It all depends. It all de- based tion. I do not know of anyone else who [has] pends on how likely each morphological change hap- successfully tried to make such a simulation ei- pens to be. If cells in their initial light-sensitive patch ther. But we are currently working on it. To must discover their appointed role by chance, all es- make it behave like real evolution is not a sim- timates of the time required to bring about just the ple task. At present our model does produce transformations their theory demands-invagina- eyes gradually on the screen, but it does not tion, aperture construction, and lens formation- look pretty, and the genetic algorithms need a will increase by orders of magnitude. fair amount of work before the model will be If Darwin were restored to pride of place in Nils- useful. But we are working on it, and it looks son and Pelger's work, the brief moment involved in both promising and exciting. their story would stretch on and on and on. These are explicit words, and they are the words FINALLY, THERE is the matter of Nilsson and of the paper's senior author. I urge readers to keep F' Pelger's computer simulation, in many ways them in mind as we return to the luckless physicist the gravamen of my complaints and the dessert of Matt Young. In my COMMENTARY essay of last this discussion. December, I quoted these remarks by Mr. Young: A computer simulation of an evolutionary pro- Creationists used to argue that ... there was cess is not a mysterious matter. A theory is given, not enough time for an eye to develop. A com- most often in ordinary mathematical language. puter simulation by Dan-Erik Nilsson and Su- The theory's elements are then mapped to ele- sanne Pelger gave the lie to that claim. ments that a computer can recognize, and its dy- but as I have namical laws, or laws of change, are replicated at a These, too, are forthright words, distance by a program. When the computer has just shown, they are false: Nilsson and Pelger's run the program, it has simulated the theory. paper contains no computer simulation, and no Although easy to grasp as a concept, a computer computer simulation has been forthcoming from simulation must meet certain nontrivial require- them in all the years since its initial publication. ments. The computer is a harsh taskmaster, and Sheer carelessness, perhaps? But now, in respond- programming demands a degree of specificity not ing to my COMMENTARY article, Matt Young has ordinarily required of a mathematical theory. The redoubled his misreading and proportionately aug- great virtue of a computer simulation is that if the mented his indignation. The full text of his re- set of objects is large, and the probability distribu- marks appears in last month's COMMENTARY; here tion and fitness function complicated, the comput- are the relevant passages: er is capable of illustrating the implications of the In describing the paper by Nilsson and Pelger ... , theory in a way that would be impossible using or- I wrote that they had performed a computer simu- dinary methods of calculation. "Hand calculations lation of the development of the eye. I did not may be sufficient for very simple models," as write, as Mr. Berlinski suggests, that they used Robert E. Keen and James Spain write in their nothing more than random variation and nat- standard text, Computer Simulation in Biology (1992), ural selection, and I know of no reference that "but computer simulation is almost essential for says they did.

[34] A SCIENTIFIC SCANDAL

·. .The paper by Nilsson and Pelger is a so- on a computer console while trying to guess at the phisticated simulation that even includes number of steps involved in transforming a light- quantum noise; it is not, contrary to Mr. sensitive patch into a fully functioning eyeball, Berlinski's assertion, a back-of-the-envelope their work could also be represented as computer- calculation. It begins with a flat, light-sensi- aided. tive patch, which they allow to become con- cave in increments of 1 percent, calculating 7 /4ATTYOUNG is hardly alone in his lavish mis- the visual acuity along the way. When some MlVV readings. The mathematician Ian Stewart, other mechanism will improve acuity faster, who should certainly know better, has made virtu- they allow, at various stages, the formation of ally the same patently false claims in Nature's Num- a graded-index lens and an iris, and then opti- bers (1995). So have many other prominent fig- mize the focus. Unless Nilsson and Pelger ures.3 But misreadings are one thing, misrepresen- performed the calculations in closed form or tations another. More than anyone else, by hand, theirs was, as I wrote, it has been a "computer Richard Dawkins simulation." Computer-aidedsimulation might who has been responsible for ac- have been a slightly better description, but not tively misrepresenting Nilsson and Pelger's work, enough to justify Mr. Berlinski's sarcasm at my and for disseminating worldwide the notion that it expense.... offers a triumphant vindication of Darwinian prin- ciples. And here is my familiar refrain: there is no simu- In a chapter of his 1995 book, River Out of Eden, lation, "sophisticated" or otherwise, in Nilsson and Dawkins writes warmly and at length about Nils- Pelger's paper, and their work rests on no such sim- son and Pelger's research. 4 Here is what he says ulation; on this point, Nilsson and I are in com- (emphasis added throughout): plete agreement. Moreover, Nilsson and Pelger do not calculate the visual acuity of any structure, and [Their] task was to set up computer models of evolving certainly not over the full 1,829 steps of their se- eyes to answer two questions ... [:] is there a smooth gradient of change, from flat quence. They suggest that various calculations have skin to full camera eye, such that every inter- been made, but they do not show how they were mediate is an improvement? . .. [and] how made or tell us where they might be found. At the long would the necessary quantity of evolu- very best, they have made such calculations for a tionary change take? handful of data points, and then joined those points In their computer by a continuous curve. models, Nilsson and Pelger made no attempt to simulate the internal There are two equations in Nilsson and Pelger's workings of cells. paper, and neither requires a computer for its solu- tion; and there are no others. Using procedures very . .. Nilsson and Pelger began with a flat much like Nilsson and Pelger's own, Mr. Young has retina atop a flat pigment layer and surmount- ed by nevertheless deduced the existence of a missing a flat, protective transparent layer. The transparent layer was allowed computer simulation on theoretical to undergo local- grounds: "Un- ized random mutations of its refractive less index. Nilsson and Pelger performed the calculations They then let the model transform itself at ran- in closed form or by hand, theirs was, as I wrote, a dom, constrained only by the requirement that computer simulation." But another possibility at any change must be small and must be an im- once suggests itself: that Nilsson and Pelger did provement on what went before. not require a computer simulation to undertake The results were swift their calculations and decisive. A trajec- because they made no such cal- tory of steadily mounting acuity led unhesitat- culations, their figure of 1,829 steps representing ingly from the flat beginning through a shal- an overall guess based on the known optical char- low indentation to a steadily deepening cup, as acteristics of existing aquatic eyes. the shape of the model eye deformed itself on the Whatever the truth-and I do not know it-Mr. computer screen.... And then, almost like a con- Young's inference is pointless. One judges a paper juring trick, a portion of this transparent filling by what it contains and one trusts an author by what he says. No doubt Matt Young is correct to 3 Among those who, by contrast, have raised (on the Internet) points observe that "computer-aided similar to my own, I would single out especially Brian Harper, a simulation might professor of mechanical engineering at Ohio State University. have been a better description" of Nilsson and Pel- 4 A version of the same material by Dawkins, "Where D'you Get ger's work. I suppose one could say that had Dan- Those Peepers," was published in the New Statesman (July 16, Erik Nilsson and Susanne Pelger rested their heads 1995).

[35] COMMENTARY APRIL 2003

condensed into a local, spherical region of And yet again how very remarkable-inasmuch higher refractive index. as Nilsson and Pelger's "computer-simulation .. This ratio is called Mattiessen's ratio. model" did not home in unerringly on Mattiessen's Nilsson and Pelger's computer-simulation model ratio, Nilsson and Pelger having done all the hom- homed in unerringly on Mattiessen's ratio. ing themselves and thus sparing their model the trouble. How very remarkable all this is-inasmuch as Each and every one of these very remarkable as- there are no computer models mentioned, cited, or the result of care- contained in Nilsson and Pelger's paper; inasmuch severations can be explained as for as Dan-Erik Nilsson denies having based his work lessness only if one first indicts their author on any computer simulations; inasmuch as Nilsson gross incompetence. and Pelger never state that their task was to "set up computer models of evolving eyes" for any reason INAL QUESTIONS. Why, in the nine years since whatsoever; inasmuch as Nilsson and Pelger as- their work appeared, have Nilsson and Pelger sume but do not prove the existence of "a smooth never dissociated themselves from claims about their gradient of change, from flat skin to full camera work that they know are unfounded? This may not eye, such that every intermediate is an improve- exactly be dishonest, but it hardly elicits admiration. ment"; and inasmuch as the original light-sensitive More seriously, what of the various masters of in- patch in Nilsson and Pelger's paper was never al- dignation, those who are usually so quick to de- lowed to undergo "localized random mutations of nounce critics of Darwin's theory as carrying out the its refractive index." devil's work? Eugenie Scott, Barbara Forrest, And how very remarkable again-inasmuch as Lawrence Krauss, Robert T Pennock, Philip Kitch- there are no computer "screens" mentioned or er, Kelly Smith, Daniel Dennett, Paul Gross, Ken cited by Nilsson and Pelger, no indication that Miller, Steven Pinker-they are all warm from com- their illustrations were computer-generated, and bat. Why have they never found reason to bring up no evidence that they ever provided anyone with a the matter of the mammalian eye and the computer real-time simulation of their paper where one simulation that does not exist? could observe, "almost like a conjuring trick," the And what should we call such a state of affairs? I "swift and decisive" results of a process that they suggest that scientific fraud will do as well as any also happen to have designed. other term.

[36] CONTROVERSY

A Scientific Scandal? David Berlinski & Critics

DAN-E. NILSSON: we do not explain how we arrive at lution" (visual acuity) becomes "sen- I appreciate the opportunity to a sum of 1,829 steps of 1 percent, sitivity of vision." He does not dis- respond to David Berlinski's essay but explanations for both are given tinguish between selection and on the 1994 paper I authored with on page 56 of our paper. He further intensity of selection. He is obvi- Susanne Pelger called "A Pes- claims that we fail to explain how ously confused by the difference be- simistic Estimate of the Time Re- morphological change relates to tween the 1-percent steps that we quired for an Eye to Evolve" ["A improvements in visual acuity, use as a unit of measure for mor- Scientific Scandal," April]. Because though pages 54 through 56 (to- phological change and the 0.005- it gives them credibility, I generally gether with the graphs and legends percent change per generation that do not debate pseudo-scientists, but in figures 1 and 3) deal with exactly is our conservative estimate of evo- I have decided to make an exception that, and in great detail. lutionary rate. here. For the rest of his essay Mr. Mr. Berlinski attempts a peculiar Apart from a mix-up in chronol- Berlinski focuses on issues where he probability argument involving the ogy and some other minor pecu- believes he has detected logical random substitution into the word liarities, the only major flaw in Mr. flaws. He is not right in a single "at" of letters "fished from an urn," Berlinski's description of our paper case, and instead reveals an insuffi- but he does not realize that his ex- is his misunderstanding of the re- cient background in visual optics, ample implies a single individual in sponse variable R, which he calls a sampling theory, basic evolutionary the population, in which case there measure of "visual acuity." It is not, theory, and more. Nor does he seem can of course be no selection at all. and the original paper does not say to have read key references such as Again, he badly needs to read Fal- so. This is his first serious mis- Warrant & McIntyre (1993), Fal- coner's standard work. take-and it gets worse. coner (1989), or Futuyma (1986). Contrary to Mr. Berlinski's claim, Mr. Berlinski's next move is to list Without such knowledge it would we calculate the spatial resolution the important information he claims be hard to grasp the details of our (visual acuity) for all parts of our is missing in our paper. (At regular paper, but it is standard scientific eye-evolution sequence, and the re- intervals he repeats the phrase: practice not to repeat lengthy rea- sults are displayed in figure 1 of our "they do not say.") But all the nec- soning when a short reference can paper. The underlying theory is ex- essary information is there. I cannot be given. plained in the main text, including reply individually to every point But there is more. Mr. Berlinski the important equation 1 and a ref- here, but two examples will do. Mr. has a problem with definitions. erence to Warrant & McIntyre Berlinski claims that there is no unit "Morphological change" becomes (1993), where this theory is derived. for morphological change and that "biological change." "Spatial reso- Yet Mr. Berlinski insists that "Nils-

[10] COMMENTARY JULY-AUGUST 2003 son and Pelger do not calculate the latest article. I had written earlier But Darwin's fears are ancient his- visual acuity of any structure." It that his disparagements of Darwin- tory: Darwin was still haunted by would be much simpler for Mr. ism are old and naive, refuted in the Paley's 1802 version of the argu- Berlinski if he went just a tiny step literature. Responding in the March ment from design. A century and a farther and denied the existence of issue ("Darwinism versus Intelligent half have passed. our paper altogether. Design"), he dismissed this airily as In the 21st century there is no Had these and other points been an unanswerable gripe. But it is not question that eyes, endlessly varied unfortunate misunderstandings, I a gripe. Nor was it meant to be an- in structure and quality, have would have been only too happy to swered in COMMENTARY. It is just a evolved. Most of the intermediates help, but I have the distinct impres- fact about the scientific literature. between a primitive patch of pho- sion that they are deliberate attempts Any reader can check for himself. tosensitive cells and the camera eye to eliminate uncomfortable scientif- Examples include: Mark Ridley, Evo- of a fish or a mammal exist. Many ic results. Why does Mr. Berlinski lution, 2nd Edition (1996); John Ger- more have existed in the past, dur- not read up on the necessary scien- hart and Marc Kirschner, Cells, Em- ing the 540 million years since there tific background? Why does he so bryos, and Evolution (1997); Rudolf A. have been eyes. blatantly misquote our paper? Why Raff, The Shape of Life (1996). So what is the fuss about? In their has he never asked me for the details Only once, in the eleven years 1994 theoretical paper, Nilsson and of the calculation he claims to want since the start of their anti-evolu- Pelger modeled one possible evolu- so badly? It is simply impossible to tion PR-blitz, have any arguments tionary pathway to the geometry of take Mr. Berlinski seriously. of Mr. Berlinski's colleagues at the a fish-like eye from a patch of pho- Mr. Berlinski is right on one Discovery Institute's Center for Sci- toresponsive cells. There were al- point only: the paper I wrote with ence and Culture appeared in the ready such cells on Earth a billion Pelger has been incorrectly cited as primary literature. That was an ear- years before there were eyes. Nils- containing a computer simulation ly philosophical monograph by the son and Pelger used pessimistic es- of eye evolution. I have not consid- Christian apologist William Demb- timates of such relevant parameters ered this to be a very serious prob- ski. Mr. Berlinski (in "Has Darwin as the intensity of selection for their lem, because a simulation would be Met His Match?," December 2002) number-crunching. The point was a mere automation of the logic in now rejects that argument as ap- to determine how many plausible, our paper. A complete simulation is plied to biology, although he gave populational micro-steps of varia- thus of moderate scientific interest, Dembski's book a glowing blurb. tion would be needed for very weak although it would be useful from an The rest of their anti-evolution selection to yield a fish-like eye- educational point of view. kvetching has been in trade books and then under reasonable assump- Our paper remains scientifically mainly from religious publishers, in tions to convert micro-steps into sound, and has not been challenged nonscientific journals, testimony to generations and years. The answer in any peer-reviewed scientific jour- legislators, interviews, speeches, and was about 350,000-a geological nal. I do not intend to take any fur- rallies for the faithful. For this, Mr. blink of the eye. This answer is just ther part in a meaningless debate Berlinski offered the crank excuse: one of many to the failed 19th-cen- with David Berlinski. scientific prejudice. And as coup de tury complaint of insufficient time Lund University main, he quoted lines from a 1986 for evolution to have taken place. Lund, Sweden essay of mine. But the burden of Mr. Berlinski misunderstands or that essay is precisely the opposite misinterprets critical elements of PAUL R. GRoss: of Mr. Berlinski's reason for quot- the paper. Then he quibbles pon- "A Scientific Scandal" is itself a ing it. It was about a distinguished derously about terms and assump- scientific scandal: the continued regular contributor to the scientific tions-and about a popular gloss of publication, in a political-cultural literature. the paper by Richard Dawkins. He opinion journal, of David Berlin- The obvious purpose of "A Sci- accuses some of his critics of fraud ski's uninformed bellyaching about entific Scandal," like Mr. Berlinski's for having failed to denounce evolutionary biology. COMMEN- other adventures in evolutionary Dawkins's use in a trade book of TARY is not the place for quasi-tech- thought, is to belittle Darwinism. certain of those terms. Mr. Berlin- nical arguments against Darwinism, He cites Darwin himself, who wor- ski's arguments are quibbles. or for reprinting the scientific pa- ried a little that his theory might not But these quibbles are beside the pers or textbook chapters that dis- be able to account for the eye. real point, which is that we lack prove them. Mr. Berlinski's real case is that Dar- grounds for believing that eyes Mr. Berlinski has several times win's fears were justified: evolu- evolved. That is false. Eyes, like found fault with me. The method is tionary theory cannot explain the anything else, could have been in- characteristic, and it is salient in this eye, and there has been a cover-up. vented at a stroke by a supernatural d

[12] COMMENTARY JULY-AUGUST 2003

designer. But there is no evidence numbers, Nilsson and Pelger calcu- I will not respond to Mr. Berlin- of it. Neither can it ever be dis- lated an average change of 0.005 per- ski's disdainful tone, nor to the proved. The only explanation, how- cent per generation. The relative cheap shots directed at me person- ever, that we have for the structure change in n generations is therefore ally. Nor will I continue the point- of eyes-as solid as any explanation (1.00005)n, which they set equal to the less distraction of whether Nilsson in science-is Darwinian evolution. overall change of morphology in their and Pelger performed a simulation Like the intelligent-design group simulation (1.011,829, where 1,829 is or a calculation. I am, however, con- as a whole, Mr. Berlinski seems un- the number of 1-percent steps re- cerned with Mr. Berlinski's con- able or unwilling to understand the quired to form an eye). The number tention that reputable scientists have newest branch of biology: evolu- 1.00005 is not, contrary to Mr. Berlin- conspired to support a technical pa- tionary developmental biology. There, ski, a percentage; it is the relative per that he finds "unfounded"; with the discovery of the develop- change of a given parameter in a sin- charging specific individuals with mental regulatory genes, we have gle generation. Nilsson and Pelger con- "fraud" is not to be taken lightly. learned how subtle, how versatile, cluded that an eye could have evolved The paper has survived peer review, and yet how simple the mechanisms in approximately 350,000 years. and has not been shown to be un- can be for transforming one biolog- Does anyone claim that an eye founded in any peer-reviewed jour- ical structure to another. (A profes- evolved precisely as Nilsson and nal. If Mr. Berlinski thinks the pa- sional but accessible account can be Pelger's simulation suggests? No. per is unfounded, let him submit a found in From DNA to Diversity: But I stand by my statement that paper of his own to a peer-reviewed Molecular Genetics and the Evolution of they have given the lie to the cre- journal and find out what the sci- Animal Design [2001] by Sean B. ationists' claim and firmly made the entific community thinks of his Carroll, Jennifer K. Grenier, and case that an eye could have evolved ideas. It is unlikely that scientific Scott D. Weatherbee. A popular but within a geologically short time. journals, which have occasionally sound insight is available in: "Which Mr. Berlinski argues, for exam- published papers on homeopathic Came First, the Feather or the ple, that morphological changes of medicine and the Bible codes, Bird?" by Richard O. Prum and Alan the skull might slow the process. would reject Mr. Berlinski's paper H. Brush, Scientific American, March Never mind that only vertebrates out of sheer prejudice. 2003.) A reader whose view of sci- have skulls, and Nilsson and Pelger's Colorado School ofMines ence comes only from Mr. Berlins- eye is, again contrary to Mr. Berlin- Golden, Colorado ki will never know of such things. ski, an invertebrate eye. The devel- JamaicaPlain, Massachusetts opment of an eye will require not MARK PERAKH: only morphological changes but It is funny that COMMENTARY- MATT YOUNG: also advancements to the nervous by no measure a scientific publica- Creationists often claim, without system and the brain. Will these re- tion-has allocated so much space presenting evidence, that there has quirements bring evolution to a to recent articles by David Berlinski. not been enough time for a complex halt? Georges Cuvier asked the With no record of scientific re- organ such as the eye to have same question in 1812, and the an- search, either in biology or in com- evolved. To examine that claim em- swer is, "no." We now know that puter science, he sets out to pro- pirically, Nilsson and Pelger devised evolution progresses in a modular nounce judgment on topics within a scenario in which an eye could have way, with different systems evolving these two fields. evolved through stages that are in parallel and nearly independent- But contrary to Mr. Berlinski's known to exist in the animal king- ly. If Mr. Berlinski thinks that vari- rhetoric, any scandal related to Nils- dom. I described their scenario in my ous modules could not have co- son and Pelger's paper occurred last letter to COMMENTARY (March), evolved, he needs to support his only in Mr. Berlinski's imagination. and David Berlinski has it almost argument quantitatively, not just Nilsson and Pelger estimate the right in "A Scientific Scandal." proclaim it. Nilsson and Pelger have time necessary for the development Briefly, Nilsson and Pelger formed shown precisely what they set out to of an eye, a calculation that entails an eye by changing various parame- show: that an eye could have certain assumptions but which is ters, such as aperture diameter, in evolved in a geologically short time viewed by many scientists as suffi- 1-percent increments, until no im- and that the eye itself is not a limit- ciently sound. (According to the Sci- provement could be made. One per- ing factor. Mr. Berlinski holds ence Citation Index, Nilsson and cent is an arbitrary number (any small against them that they did not per- Pelger's article has been positively increment will suffice) and does not form the full-fledged simulation he referenced in at least 25 peer- represent the change in a single gen- wants them to have done and seems reviewed scientific publications.) eration. Using what they and Richard to think that their calculation is But Mr. Berlinski, unlike all these Dawkins describe as conservative therefore somehow faulty. scientists._ , does not like Nilsson and I

[14] COMMENTARY JULY-AUGUST 2003

Pelger's conclusion, and obfuscates bringing clarity to difficult scientific Nilsson and Pelger's paper as a the issue by discussing the distinc- issues. If he did, he would not have stochastic computer simulation of tions among computer simulations, made so many misrepresentations in the evolution of the eye (it was actu- models, and calculations. These se- describing Nilsson and Pelger's ally a mathematical model). But Mr. mantic exercises are inconsequen- work. Two examples: Mr. Berlinski's Berlinski should remove the plank tial to the real question: whether an claim that their model eyes were from his own (discussion of the) eye. eye could have developed in a geo- simply "flogged up an adaptive peak" He asserts that one of the problems logically short time via a Darwini- ignores the fact that establishing the that Nilsson and Pelger did not con- an mechanism, as Nilsson and Pel- existence of such a peak was one of sider was how the skull would be ger and scores of biologists familiar the primary accomplishments of the "reconstructed" to include eye sock- with their work think. paper. That there is a smooth gradi- ets. But as any decent student of A reader cannot fail to notice an ent of increasing visual acuity linking even high-school biology would especially appalling feature of Mr. a light-sensitive spot to a lens-bear- know, eyes evolved before bones. Berlinski's escapade: he accuses ten ing eye is a discovery that they made, Cephalochordates, the closest inver- respected scientists of "scientific not a foregone conclusion. And his tebrate relatives of vertebrates, have fraud." The reason for that prepos- claim that "in their paper there is no primitive eyes but no bones. In fact, terous accusation is that they did mention whatsoever of randomly oc- based on genetic evidence, many bi- not repudiate Nilsson and Pelger's curring changes" falls flat, since the ologists now think that vertebrate work. Mr. Berlinski apparently can- need for such changes is explicitly eyes share a common ancestral eye- not imagine that these scientists, mentioned in the discussion section spot with insect eyes. among them professional biologists of the paper, and is plainly implied To envision the evolution of the and physicists with records of sub- throughout. eye as occurring on some kind of stantial achievement, can have an In addition, Mr. Berlinski would mythical eyeless fish with a fully opinion of Nilsson and Pelger's not have unloaded so many spuri- formed skull and brain is a typical work different from his own. His ous criticisms. For example, his creationist straw man. Biologists accusation sounds even odder com- query-"why is selection pressure know that all manner of gradations ing from a man who provided rave held constant over the course of of eye complexity exist in extant or- blurbs for books by William Demb- 300,000 years"-is easily answered ganisms, from creatures with a sin- ski and Michael Behe even though, by noting that it was held constant gle photoreceptor cell, through the as is clear from his article in the De- at a value that was ludicrously low various stages that Nilsson and Pel- cember 2002 COMMENTARY, he is for almost any environment. ger depict, to the advanced camera- actually in disagreement with them Once we have swept the field of eyes of mammals and cephalopods. regarding essential parts of their as- Mr. Berlinski's distortions we are left Sometimes the whole sequence from sertions. Maybe by his standards with a few simple facts. (1) Several eyespot to advanced eye with lens this is a manifestation of integrity, decades of research on the evolution can be seen in a single group (e.g., but to me it looks more like an ex- of eyes has not only made it plain snails), yet another thing Mr. Berlin- pediency whose roots are not exact- that eyes have evolved, but has also ski would have known had he fol- ly in the search for scientific truth. revealed the major steps through lowed Nilsson and Pelger's reference Bonsall, California which they did so. (2) Nilsson and to the classic work on eye evolution, Pelger's paper provides an elegant a 56-page article by Salvini-Plawen JASON ROSENHOUSE: capstone for this research, by pro- and Mayr in Evolutionary Biology Connoisseurs of pseudoscience viding a convincing calculation for (vol. 10, 1977) called "On the Evo- will recognize in David Berlinski's an upper limit on the time required lution of Photoreceptors and Eyes." latest essay the standard tropes of for an eye to evolve. (3) Minor er- That paper answers many of the the crank's playbook: the smug sar- rors in popular treatments of Nils- questions that Mr. Berlinski asserts castic tone, the barrage of bullet- son and Pelger's paper do nothing to are unanswered or unanswerable. point criticisms to create the illu- change facts (1) and (2). (4) Finally, If Mr. Berlinski is going to de- sion that something truly rotten is David Berlinski is not a reliable clare as bunk the central organizing being exposed (criticisms he knows source for scientific information. theory of biology, he should take the will be answered by nothing more Kansas State University matter up with biologists in the pro- formidable than a few indignant let- Manhattan, Kansas fessional literature rather than in fo- ters), the crude baiting of scholars rums like COMMENTARY, wherein of vastly greater accomplishment NICK MATZKE: elementary questions like "which than he, and the presentation of mi- David Berlinski should be con- came first, skulls or eyes?" can be nor errors as tantamount to fraud. gratulated for pointing out Richard botched and yet still be published. Goleta. California Mr. Berlinski has no interest in Dawkins's...... inaccurate descriptionI-.... of

[16] COMMENTARY JULY-AUGUST 2003

DAVID SAFIR: David Berlinski Gross, Jason Rosenhouse, Matt Once again, David Berlinski has Young, and Mark Perakh have shown how a truly scientific inquiry In "A Scientific Scandal," I ob- replied to previous essays of mine in can expose academic and intellectu- served that Dan-E. Nilsson and Su- COMMENTARY-I hoped that self- al fraud by evolutionists. As a physi- sanne Pelger's paper, "A Pessimistic interest, if nothing else, might have cian, I have always been made uneasy Estimate of the Time Required for prompted a moment of critical self- by the assertions offered by propo- an Eye to Evolve," was a critic's reflection. No very delicate moral nents of evolution to explain com- smorgasbord. There are so many sense is involved in determining that plex biological life. Mr. Berlinski things wrong with it that even the fraud is fraud. If Richard Dawkins is shows exactly how the process works: finickiest of eaters could leave the one of their own, all the more rea- start with the belief that no other table well-satisfied and ready for a son to apply to him the moral stan- possible explanation for the diversi- round of Alka-Seltzer. But, in itself, dards that Messrs. Gross, Rosen- ty of life on earth could exist other there is nothing here that suggests a house, Young, and Perakh are than what we think we know about scandal. Dan-E. Nilsson is a distin- accustomed to applying to their in- evolution; demonstrate utter con- guished scientist. Witness his discov- tellectual enemies. tempt for other ideas (ad-hominem ery that the mysid shrimp, Dioptro- Reading their letters, I realize attacks are often employed here); mysis pauciponisa, is an organism that they had no intention of say- then simply invent a pathway de- whose eyes are at once simple and ing boo. What could I have been scribing how it might have been pos- compound (D. Nilsson, R.E Modlin, thinking? sible to get from point A to point B- "A Mysid Shrimp Carrying a Pair of from a light-sensitive spot, say, to a Binoculars," Journalof Experimental DAN-E. NILSSON is persuaded that complex eye. Where I come from Biology, Vol. 189, pp. 213-2 36 , 1994), I wrote my essay because I am this is called nonsense. or his precise work on the optical sys- moved to reject "uncomfortable sci- I would feel better about a the- tem of the butterfly (D. NilSson, entific results.' He is mistaken. The orist like Richard Dawkins if he did M.E Land,J. Howard, "Optics of the length of time required to form an not pontificate about how glori- Butterfly Eye," Journal of Compara- eye is a matter of perfect indiffer- ously perfect his explanations are. tive Physiology, A 162, 341-366, 1988). ence to me; had he and Susanne I cast my fate instead with scien- Together with Susanne Pelger, he has Pelger been able to demonstrate tists like Mr. Berlinski who keep an simply written a silly paper. It hap- that the eye was in fact formed over open mind. The jury is still out, af- pens. And in the literature of evolu- the course of a long weekend in the ter all, ad will be for a very long tionary biology, it happens very often. Hamptons, I would have warmly time. No, the scientific scandal lies congratulated them. As I have many Los Gatos, California elsewhere. Nilsson and Pelger's pa- times remarked, I have no creation- per has gained currency in both the ist agenda whatsoever and, beyond NORMAN P. GENTIEU: popular and the scientific press be- respecting the injunction to have a As a retired science writer, I ap- cause it has been misrepresented as good time all the time, no religious preciated David Berlinski's superb a computer simulation, most no- principles, either. Evolution long, analysis refuting Nilsson and Pel- tably by Richard Dawkins. Word evolution short-it is all the same ger's simplistic scenario of the evo- spread from Dawkins's mouth to any to me. I criticized their work not be- lution of the mammalian eye. To number of eagerly cupped but woe- cause its conclusions are unwelcome account for the perfection of that fully gullible ears. Subsequent ref- but because they are absurd. incredibly complex organ by means erences to Nilsson and Pelger's work The vertebrate eye, Nilsson and of formulaic fumblings is nothing have ignored what they actually Pelger claim, emerged from a patch less than preposterous. I wonder if wrote in favor of that missing com- of light-sensitive cells. Climbing up Nilsson and Pelger might some day puter simulation, in a nice example evolution's greasy pole, or adaptive use this iffy method to explain the of a virtual form of virtual reality fi- peak, those cells got to where they development of stereoscopic color nally displacing the real thing alto- were going by invagination, aper- vision. gether. This misrepresentation of ture constriction, and lens forma- "A Scientific Scandal" is an apt scientific work is a species of fraud, tion. In explaining the evolution of name for the docile acceptance of a no different in kind from plagiarism the eye in terms of such global ge- dubious theory. What has happened in journalism or the fabrication of ometrical processes, Nilsson and to vetting? Back in the 1950's, the data in experimental physics. It is the Pelger rather resemble an art histo- science establishment did not hesi- indifference to this fraud that I de- rian prepared to explain the emer- tate to zap Immanuel Velikovsky nounced as scandalous. gence of the Mona Lisa in terms of and his Worlds in Collision. Recognizing so many fond famil- preparing the wood, mixing the Philadelhia.Pennsylvania iar faces among my critics-Paul paint. and filling in the details. The Ir- - -.._ ..... _' -d...... r--)-a--

[18] COMMENTARY JULY-AUGUST 2003

conclusion-that Leonardo com- proved in just the way that Nilsson Pelger's work is devoted to the evo- pleted his masterpiece in more than and Pelger's graph indicates? lution of the camera eye character- a minute and less than a lifetime- Equation 1 is of scant help in this istic of fish and cephalopods. The- while based squarely on the facts, regard, despite Nilsson's insistence oretical considerations that apply to seems rather less than a contribu- that it is important. Drawing a con- bugs do not necessarily apply to fish tion to understanding. nection among visual acuity, focal or octopuses, the more so since It is hardly surprising, then, that length, light intensity, and noise, the their eyes are structurally different, while theoretical optics serves qual- equation specifies the local maxi- as are their evolutionary histories. itatively to justify the overall con- mum of a curve, the place where it Writing about the compound eye, nection Nilsson and Pelger draw stops rising. In other words, it spec- Nilsson himself has remarked that between morphology and visual ifies a point; and it does nothing "it is only a small exaggeration to acuity, nothing in their paper and more. "We can now use this rela- say that evolution seems to be fight- nothing in their references justifies tionship," Nilsson and Pelger nev- ing a desperate battle to improve a the quantitative relationships they ertheless declare, "to plot resolution basically disastrous design" (Dan-E. employ to reach their quantitative against aperture diameter." They Nilsson, "Optics and Evolution of conclusion. To be sure, Mr. Nilsson can do nothing of the sort, at least the Compound Eye," in Facets of Vi- denies that this is so. "Contrary to not in my calculus class. Knowing sion, edited by D.G. Stavenga & Mr. Berlinski's claim," he writes, that a man has reached the summit R.C. Hardie, p. 3075, 1989). What- of Mt. Everest, we still know noth- ever the desperate battle going on we calculate the spatial resolution ing about the route he has taken to among the arthropods, there is no (visual acuity) for all parts of our get there. What is needed if Nils- battle at all taking place among the eye-evolution sequence, and the son and Pelger are to justify their vertebrates or the cephalopods. results are displayed in figure 1 graph is the equation from which Nilsson and Pelger's eye moves of our paper. The underlying equation 1 has been from triumph to triumph with theory is explained in the main derived by dif- serene and text, including the important ferentiation. It is not there, just remarkable celerity. equation 1 and a reference to where I said it would not be. If the papers by Snyder and War- Warrant & McIntyre (1993), Similarly with Nilsson and Pel- rant & McIntyre say nothing about where this theory is derived. ger's references, which do nothing fish or octopuses, neither do they to support their argument. Quite say anything about evolution. No In fact, no underlying theory what- the contrary. Three papers are at is- mention there of Darwin's theory, soever is explained in Nilsson and sue: (1) A.W. Snyder, S. Laughlin, no discussion of morphology, not a Pelger's main text, or in the legend and D. Stavenga, "Information Ca- word about invagination, aperture to figure 1; and while they do assert pacity of the Eyes" (Vision Research, constriction, or lens formation, and that calculations were made, they do vol. 17, 1163-1175, 1977); (2) A.W. nothing about the time required to not say where they were made or Snyder, "Physics of Vision in Com- form an eye, whether simple, com- how they were carried out. The pound Eyes" (in Vision in Inverte- pound, or camera-like. burden of Mr. Nilsson's denials is brates, Handbook of Sensory Phys- The purpose of these three pa- conveyed entirely by equation 1 and iology, edited by H. Autrum, vol. pers is otherwise. No less than any by his references. VII/6A, pp. 225-313, 1979); and (3) other system of communication, the Let us start with equation 1, and E. J. Warrant & P.D. McIntyre, eye represents a balance struck be- with figure lb that this equation is "Arthropod Eye Design and the tween signal and noise. There is the said to control. It is in figure lb Physical Limits to Spatial Resolving object out there in the real world- that aperture constriction takes Power" (Progressin Neurobiology, vol. whether a point source like a star, or over from invagination in getting 40, pp. 413-461, 1993). Of these pa- an extended source like a grating of an imaginary eye to see better. The pers, the first is recapitulated (and light and dark lines-and there is its graph juxtaposes aperture size corrected) in the second, and the image trembling on the tips of the against detectable spatial resolution. second is summarized in the third. retina's budded nerve cells. Slippage Having dimpled itself in figure 1a, In what follows, references to Sny- arises between what the object is Nilsson and Pelger's blob is now der are always to the Snyder of his and how it is seen. Noise occurs in busy puckering its topmost surface second paper. the visual system as the result of the to form a pinhole in figure lb.* In As their titles might suggest, both random nature of photon emission, a general way, the curve they pre- "Physics of Vision in Compound and it also occurs as the result of in- sent is unremarkable. No one Eyes" and "Arthropod Eye Design herent imperfections in the eye's op- doubts that spatial resolution is im- and the Physical Limits to Spatial * Three curves are given in figure b, repre- proved in an eye when its aperture Resolving Power" deal with com- senting three different levels of light inten- is constricted. But why is it im- pound invertebrate eves. Nilsson and sity, but this plays no role in what follows. I

[20] CONTROVERSY tical system. The theoretical opti- optical system depends on a host of ness. Determining the shape of the cian abbreviates these limitations in factors, including head size and eye angular-sensitivity function is a lit- one mathematical instrument. movement, most of which Nilsson tle like trying to guess an astronaut's Imagine one of Nilsson and Pel- and Pelger ignore. Nor, for that weight in space. Scales are not like- ger's plucky light-sensitive cells, and matter, do these equations taken in ly to be of use. In an early paper then extend two flanking lines from isolation justify any particular quan- dealing with this subject and devot- the cell up past the constricted aper- titative conclusions. Until the angu- ed experimentally to flies, K.G. ture and out into space, so that the lar-sensitivity function is specified, G6tz noted that the angular-sensi- cell and those two flanking lines whether theoretically or experimen- tivity function in Drosophilaseemed form a cone with a flat top. In the tally, its role is ceremonial. to follow what is known mathemat- center of the cone, where a cherry Such specification is no easy busi- ically as a Gaussian probability dis- would sit atop the ice cream, there is a light source. The cherry moves to the sides of the cone in angular steps; the cell dutifully responds. The correlation between moving cherry and twitching cell constitutes the optician's "angular-sensitivity function." Equation B15 (p. 238) in Snyder's "Physics of Vision in Compound Eyes" defines the signal-to-noise ra- tio of a hypothetical eye in terms of noise, modulation contrast (the dif- ference in intensity between black and white stripes in a grating), and the modulation-transfer function, which is simply a mathematical transformation of the eye's angular- sensitivity function (its Fourier transform). Lumbering in Snyder's footsteps, Warrant & McIntyre split his equation into two of their own (equations 10 and 11 in Warrant & McIntyre, p. 430), the one describ- ing the signal, the other the noise in a hypothetical visual system. They observe what is in any case obvious: whatever the parameters affecting visual acuity, signal and noise will always reach a point where the first is drowned out by the second and the system fails, a point evident enough to anyone trying to see in the dark. These equations lead by primo- geniture to Nilsson and Pelger's equation 1, which, as it happens, does not appear anywhere in their sources in the form in which they express it. But neither Snyder's orig- inal equation nor Warrant & McIn- tyre's bright bursting clones in any way suggest that the tipping point between signal and noise is unique. The ratio of signal to noise in an

[21] COMMENTARY JULY-AUGUST 2003

tribution (K.G. G6tz, "Die optis- But theory is one thing, and liv- a theory describing the theoretical chen Ubertragungseigenschaften ing flesh another. Staking their all optics of compound invertebrate der Komplexaugen von Drosophila," on Snyder's model, Nilsson and Pel- eyes; (4) failed to explain why his Kybernetik, 2, pp. 215-221, 1965). It ger must live with its consequences. own work has neglected to specify was an interesting idea, but one that "Having considered the physical any relevant biological parameter led to very considerable computa- limitations to resolving power," precisely; and (5) championed his tional difficulties. Snyder wrote, "in addition to the results by means of assumptions that Looking Gbtz-ward, and under- absolute sensitivity of eyes, we now his own sources indicate are false standably recoiling, Snyder adopt- apply our concepts to real com- across a wide range of organisms. ed a different strategy. In assessing pound eyes." This is something that In acknowledgments to their pa- the weight of an astronaut in space, Nilsson and Pelger never do. And per, Nilsson & Pelger thank E. J. it is simpler to count the calories he no wonder. For Snyder then added Warrant for help with their compu- consumes and the exercise he un- the rather important caveat that tations; in the acknowledgments to dergoes than to try to measure his bringing theory to bear on life "re- their paper, Warrant & McIntyre weight directly. His weight, although quires precise knowledge [of various thank Mr. Nilsson for critically unmeasured, follows inferential- optical parameters] in the various reading what they have written. ly. In just the same way, Snyder regions of the eye" (Snyder, p. 276, Schnapps all around, I am sure. thought to consider the angular- emphasis in the original). sensitivity function indirectly by con- If precise knowledge is needed in I TURN next to the morphological sidering the structures that deter- applying Snyder's model, precise de- units that are missing from Nilsson mined its shape. These, he assumed, tail is what is lacking in Nilsson and and Pelger's paper. It makes no were the eye's retinal receptive Pelger's paper. Precise detail? Any sense to say of a ruler that it is one field-the area of the retina re- detail whatsoever. long. One what? When the "what" sponding to signals-and its optical And for obvious reasons. When has been specified, a physical unit "blur spot"-the smeared image tested, Snyder's model turns out to has been indicated: one inch, say, in represented on the retina corre- be false across a wide range of ar- the case of length, one pound in the sponding to the sharp object being thropods. As Warrant & McIntyre case of weight. If one inch and one seen. Let them both, he declared, note glumly, "The model, on the pound are units, length and weight be identically Gaussian. Why not? whole, works best for those eyes for are their dimensions. Only an ori- Both parameters had simple math- which it was originally formulated- gin in zero remains to be specified ematical natures. The retinal recep- apposition compound eyes func- to complete the picture. tive field is given as the ratio of the tioning according to geometrical In my essay, I observed that Nils- rhabdom's diameter to its posterior optics-but recent careful and sen- son and Pelger had not specified nodal distance, the optical blur as sitive measurements of angular sen- their unit of morphological change. the ratio of the wavelength of stim- sitivity reveal that even in these Nilsson now asks me to consider ulating light to the eye's aperture. types of eye, the model often per- again their remarks on p. 56 of their From this the shape of the angular- forms poorly." Readers may consult paper. There, he is certain, I will sensitivity function followed. figure 34 (p. 441) of Warrant & find the missing unit carefully ex- The result is known as the Sny- McIntyre's paper to see how poor- plained. Here is what they write, der model. "The great beauty of this ly the Snyder model does. In stud- and it is all that they write: "Our model," Warrant & McIntyre re- ies of the locust Locustia, real and principles have been to use whole- mark (in words that they have itali- predicted angular-sensitivity func- length measurements of straight cized), "is that if one knows some very tions do not even share the same structures, arc lengths of curved simple anatomicalinformation about qualitative shape. structures, and height and width of the eye"-i.e., the nature of its opti- Responding to my observation voluminous structures." cal blur spot and retinal receptive that no quantitative argument sup- Very well. These are the funda- field--"one has the ability to predict... ports their quantitative conclu- mental units. They are none too the approximate shape of the angular- sions-no argument at all, in fact- clearly explained-try estimating sensitivity function" (p. 434). In re- Mr. Nilsson has thus (1) offered a the volume of a donut by looking at ferring to Warrant & McIntyre, mathematically incoherent appeal its height and width-but I know Nilsson and Pelger are, in fact, ap- to his only equation; (2) cited refer- roughly what Nilsson and Pelger pealing to Snyder, the maztre behind ences that make no mention of any are getting at. What they do not say their masters-for, like Snyder, they, morphological or evolutionary pro- is how these three separate funda- too, assume that retinal receptive cess; (3) defended a theory intend- mental units are combined in a sin- fields and optical blur spots are ed to describe the evolution of ver- gle overall derived unit of change. identically Gaussian (p. 54). tebrate camera eves bv referrin to --1-- -J - -- A homely example may make this

[22] CONTROVERSY more vivid. Except for the fact that vious that the eye was never a real it cannot see, a Swedish meatball is threat to Darwin's theory of evolu- rather like an eye. And plainly it tion" (p. 58). By "this context," they makes no sense to ask of two Swedish mean one in which only "eye geom- meatballs, one of them twice as etry" and "optical structures" are up greasy but half as wide as the other, for grabs. But whether in this con- which of them is bigger-at least text or any other, it is as a defense of not until units of grease and length Darwin's theory that Nilsson and have been combined. But this is, in Pelger's theory fails most obviously. general, no easy task, not even when Let me review the chief steps in shape alone is under consideration. their argument. There is morpho- "It is important to keep in mind," logical change on the one hand, vi- C.P. Klingenberg and L. J. Leamy sual acuity on the other. As their write ("Quantitative Genetics of population of light-sensitive cells al- Geometric Shape in the Mouse ters its geometry-by means never Mandible," Evolution, 55(11), pp. specified-visual acuity perks up. In 2342-2352, 2001), "that shape is a all, they assert, 1,829 steps are in- multivariate feature and cannot be volved in tracing a path from their easily divided into scalar traits with- first patch to their final "product." out imposing arbitrary constraints Just how do Nilsson and Pelger's on the results of the analysis." To light-sensitive cells move from one see how difficult a conceptual prob- step on that path to the next? I am lem Nilsson and Pelger have set not asking for the details, but for the themselves, readers may follow the odds. There are two possibilities. trail of Klingenberg & Leamy's ref- Having reached the first step on the erences to the badlands of current path, the probability that they will work on geometric morphometrics. reach the second (and so on to the Operating perhaps on the princi- last) is either one or less than one. If ple that a difficulty disclosed is a dif- one, their theory cannot be Darwini- ficulty denied, Nilsson and Pelger an-there are no random changes. If do mention this very point, citing less than one, it cannot be right- an example of their own on p. 56 to there is no way to cover 4,829 steps show just how arbitrary can be the in roughly 300,000 generations if business of calculating combined or each step must be discounted by the derived units. In then justifying probability of its occurrence. their own procedure, which is nev- Demonstrating the existence of a er explained, they remark: "As we path between two points in the his- are going to relate our measure of tory of life is in general not hard. morphological change only to gen- What is hard is determining how eral estimates of phenotypic varia- the path was discovered. (This was tion" in visual acuity, "we will be the point of the linguistic example safe as long we avoid unorthodox I offered in my essay.) If one as- and strange ways of comparing ori- sumes, as Nilsson and Pelger do, gin and product." that probabilities need not be taken Origin and product? I am sure into account because all transitions they meant origin and unit. No occur with a probability of one, matter. The remark speaks for it- there is no problem to be dis- self. cussed-but nothing of any con- ceivable interest, either. In re- THERE IS next the matter of random sponding to this obvious point by variation: the heart of the matter so generously suggesting that I need to far as I am concerned. Nilsson and spend more time by the lamp with Pelger's paper is not an exercise in D.S. Falconer's Principlesof Quanti- theoretical optics. It is intended to tative Genetics, Mr. Nilsson has cov- serve polemical purposes. Thus, ered an embarrassment by ad- "In this context it is ob- dressing an irrelevance. Neither thevd write:

[23] COMMENTARY JULY-AUGUST 2003

population size nor natural selec- tence, it excludes one of its own, a writes, they have surely wasted their tion is at issue. scientist who like L.V. Heilbrunn time. The existence of such a path A few minor matters. Falconer's has published in the literature. Such is hardly in doubt. Every normal response variable R is a measure, all men are entitled to wear the gold human being creates an eye from a right: a measure of the extent to cufflinks with the crossed golf clubs; patch of photoresponsive cells in which the mean of some quantitative keeping them out would be irre- nine months. phenotypic character-snout length, sponsible. But keeping out the oth- I certainly agree that the "only crop yield, scab color, or scrotum ers is not only good science but explanation we have for the struc- size (examples from the literature, I good sense. Ipse dixit. ture of the eye .. is Darwinian evo- am afraid)-rises or falls as the result A few other points deserve com- lution." But neither an orchestra of natural selection. Just what I said, ment. In offering Nilsson and Pel- nor an explanation becomes good just as I explained. Although I of- ger the oil of his approval, Mr. by being the only game in town. fered no definitions in my essay, the Gross affirms that I have misun- On the other hand, I disagree paraphrases I employed were harm- derstood or misinterpreted critical that Darwin's theory is as "solid as less. Why not say "sensitivity to vi- elements of their paper. In keeping any explanation in science." Dis- sion" instead of "visual acuity," just with his longstanding policy of nev- agree? I regard the claim as prepos- to vary pace and prose? But in one er documenting his discontent, he terous. Quantum electrodynamics respect, Mr. Nilsson is right: I did does not say which elements. As I is accurate to thirteen or so decimal not distinguish between selection keep reminding him, this is not places; so, too, general relativity. A and intensity of selection. Neither sporting. Still, it is inconceivably leaf trembling in the wrong way does he. Neither does Falconer's re- droll to see Mr. Gross excusing would suffice to shatter either the- sponse statistic, which contains only Richard Dawkins's misrepresenta- ory. What can Darwinian theory of- one selectional parameter, and that tion of Nilsson and Pelger's work by fer in comparison? one measuring the intensity of se- appealing to the fact that Dawkins Finally, I would hardly dispute lection. Neither does anyone else in expressed his views in a trade book. Mr. Gross's claim that "with the dis- this context. Mr. Gross apparently believes that covery of the developmental regu- His paper with Susanne Pelger, outside the country club, a man can latory genes, we have learned how Mr. Nilsson writes, has never been say anything he wants, a policy that subtle, how versatile, and yet how criticized in the peer-reviewed lit- he would not dream of applying to simple the mechanism can be for erature. I am certain that this is so. critics of Darwin's theory. transforming one biological struc- A few of Mr. Gross's remarks ture to another." If he were to re-read PAUL R. GROSS takes the occasion suggest a need for remedial reading. the correspondence (COMMENTARY, of his current letter to assure read- I have never argued that "evolu- September 1996) following the pub- ers that what he meant in his last tionary theory cannot explain the lication of my "The Deniable Dar- letter he did not say and what he eye." How on earth would I know win" (June 1996), he could not fail said he did not mean. Like golf, Mr. that? And explain what in particu- to be struck by my reply to his own Gross suggested in the 1986 essay lar? Its emergence, its structure, its letter, in which I specifically called from which I uncharitably quoted physiology, its biochemistry? What attention to work on regulatory in the March COMMENTARY, sci- I contended specifically is that Nils- genes and eye formation-the very ence is rather a clubby affair, and son and Pelger's paper is just nuts. work that he now suggests I am just as a great many men prefer to Conspiracies and cover-ups are, in keeping from my readers. Subtle cover the links sedately in the com- any case, not in my line, and I nev- and versatile, those genes? Yes, in- pany of men like themselves-tas- er suggested or supposed that evo- deed. Absolutely astonishing? That, sels on their shoes, alligators on lutionary biologists who failed to too. But hardly a triumph of Dar- their polo shirts-so scientists pre- criticize Richard Dawkins for mis- win's theory. For one thing, no fer to keep company with their own, representing Nilsson and Pelger did Darwinian theorist had predicted men and women who share their so as part of a conspiracy. Like the existence of these genes; for an- tastes, point of view, outlook on life. droshky horses, they were only do- other, no Darwinian theorist has ex- These are sentiments so candid ing what comes naturally: turning a plained their emergence. The facts that I was surprised to find Mr. blind eye. are simply far more fascinating than Gross expressing them. But he is If the burden of Ni sson and Pel- anything that poor drab Darwin, now prepared to disown what he ger's paper was to demonstrate the endlessly sifting time and chance, said. The club is just fine, and just existence of "one possible evolu- could possibly have imagined. look at those splendid greens! The tionary pathway to the geometry of admissions board is to be faulted a fish-like eye from a patch of pho- CITING THOSE ever useful but eter- only when, by accident or inadver- toresponsive cells," as Mr. Gross nally anonymous "creationists,"

[24] CONTROVERSY

Matt Young argues yet again, as he en place by quantitative steps. In the whether an eye could have devel- did in our earlier exchange, that first place, I made no such claim, if oped in a geologically short time via Nilsson and Pelger have given the only because its truth struck me as a Darwinianmechanism" (emphasis lie to creationist claims. If it was obvious. But were I to make such a added). But then, although quite their computer simulation that orig- claim I would observe, as Richard confident that I am wrong in my inally lent ardor to his asseverations, Dawkins does, that to the extent criticisms, he offers nothing by way now it is their paper itself. Mr. that simultaneous and parallel of rebuttal, Like so many of these Young is a man plainly prepared to changes are required to form a com- martial-arts types, he is too busy rely on an endless series of fallback plex organ, to that extent does the preparing himself to run from the positions. In the end, he may have hypothesis of random variation and field with honor to bother doing to argue that his refutation is its own natural selection become implausi- battle. best friend, and that Nilsson and ble. It is one thing to find a single Contrary to what Mr. Perakh as- Pelger's paper is itself superfluous. needle in a haystack, quite another serts, not only can I imagine, I do No one doubts that the eye has to find a dozen needles in a dozen not doubt, that "distinguished sci- evolved. Not me, in any event. Fish haystacks at precisely the same time. entists," many with a record of have eyes; rocks do not. Those eyes Surely the burden of proof in such "substantial achievement," can have came from somewhere-right?-and matters is not mine. I am not an opinion different from my own. if coming from somewhere counts obliged to defend such mathemati- It happens all the time. I would not as evolution, count me among its cal trivialities as the proposition that dream of accusing ten respected sci- champions. No one doubts, further- as independent events are multiplied entists of fraud simply because they more, that the "eye could have in number, their joint probability of passed on the opportunity to have a evolved in 350,000 generations." As occurrence plummets. go at Nilsson and Pilger. The men I remarked earlier, the eye could I have no idea what Mr. Young and women I criticized earned my have evolved in a weekend. The is- means when he writes that the contempt the hard and dirty way, by sue is whether it could have evolved number 1.00005 is not a percentage. saying nothing about scientific mis- in 350,000 generations given the con- Every number can be expressed as conduct when it was right under straints of random variationand nat- a percent, and every percent is a their noses. ural selection. pure number. But he gets half cred- I have absolutely no idea. Neither it for spotting a slip: the figure of LIKE MR. PERAKH and Paul R. do Nilsson and Pelger. And neither 1.00005 between parentheses on p. Gross, Jason Rosenhouse regards does Matt Young. 33 in my text should have been .005. Richard Dawkins's misrepresenta- Arguing now from the last trench Mr. Nilsson, who also spotted the tion of Nilsson and Pelger's work as before the bunker, Mr. Young writes slip, gets the other half. Me? I blame a "minor error." Some minor, some that Nilsson and Pelger's paper my editors. error. What, may I ask, is the dif- deals with the development of in- Finally, I did not fault the scien- ference between inventing data out vertebrate eyes, and triumphantly tific community for failing to criti- of whole cloth and inventing a com- chides me for overlooking this cize Nilsson and Pelger's work. I did puter simulation out of whole cloth? point. On p. 56 of their paper, Nils- the job of criticism myself. I faulted Should not evolutionary biologists son and Pelger write: "After con- the Darwinian community-Mr. be held to the same standards as striction of the aperture and the Young included-for failing to de- physicists? Or even journalists? gradual formation of a lens, the fi- nounce scientific fraud, specifically What part of the declaration that nal product becomes a focused cam- the misrepresentation of Nilsson and fraud is fraud does he fail to en- era-type eye with the geometry typ- Pelger's work by Richard Dawkins. dorse? These are not semantic is- ical for aquatic animals (e.g. fish and Now I see that Mr. Young feels I sues. If I claimed in print that Mr. cephalopods)." Fish are, of course, have manhandled him in these ex- Rosenhouse has four eyes, his de- vertebrates, as anyone who has changes. Too bad. COMMENTARY is nials would not turn on what I picked the flesh from a flounder not some academic mouse hole. meant. Two eyes, I am sure he knows. Perhaps I will be forgiven if would say, are not there. Two eyes, I refer to this exchange as shooting MARK PERAKH, a sensei of the "not- and one computer simulation. fish in a barrel. ed scientists say" school of self- Mr. Rosenhouse believes that Making the point that the emer- defense, is right in one respect: the Nilsson and Pelger- made an impor- gence of even the most modest eye computer simulation missing from tant discovery: namely, "that there will require simultaneous and par- Nilsson and Pelger's paper has no is a smooth gradient of increasing allel evolutionary development, Mr. bearing on what they actually said visual acuity linking a light-sensitive Young asks that I defend my claim and claimed. And right in a second spot to a lens-bearing eye." This is not their discovery, it is a restate- renert: "The real uestion [isl ,,J not have tak- | that this process could I ---- 1-

[25] COMMENTARY JULY-AUGUST 2003

ment of their chief assumption. that it is "ludicrously low for almost But whoever they are, I am not "The model sequence is made," any environment." Is it indeed? The among them. Quite the contrary, I they write, "such that every part of figure that Mr. Rosenhouse calls lu- am as eager to do right by the snails it, no matter how small, results in dicrous, Nilsson and Pelger term as he is: why should he think other- an increase of the spatial informa- pessimistic, and Mr. Gross reason- wise? It is only when he passes to tion the eye can detect" (p. 53). able. The correct term is arbitrary- matters of fact that we part compa- Note: made, not discovered. as in, it is anyone's guess what the ny. To repeat, the flaw in Nilsson variance among a bunch of fish Nilsson and Pelger's theory is in- and Pelger's work to which I attach might have been a couple of million tended to encompass the evolution the greatest importance is that, as a years ago. Studies of variance and of the eye in fish and cephalapods. defense of Darwinian theory, it heredity typically deal with tiny Fish indisputably have bones, an at- makes no mention of Darwinian populations and small periods of tractive skull, and for the most part principles. Those principles demand time. Studying the collard flycatch- two staring eyes. The cephalochor- that biological change be driven er, Ficedulaalbicollis, Merilla, Kruuk, date Branchiostoma (Amphioxus in a first by random variation and then and Sheldon collected eighteen now out-of-date system of nomen- by natural selection. There are no years of data for 17,171 nestlings in clature) is widely taken by paleon- random variations in Nilsson and order to reach some quite modest tologists to be a very plausible an- Pelger's theory. Whatever else their quantitative conclusions (J. Meril- cestral model to the vertebrates. It light-sensitive cells may be doing, la, L.E.B. Kruuk, and B.C. Sheldon, has certain vertebrate features while they are not throwing down dice or "Natural Selection on the Genetic lacking others. These others include flipping coins to figure out where Component of Variance in Body bones, a skull, a brain, ind paired they are going next. Condition in a Wild Bird Popula- sensory organs: in other words, it has Mr. Rosenhouse's conviction that tion," Journal of Evolutionary Biolo- no eyes. Mr. Matzke's very confi- the randomly occurring changes re- gy 14, pp. 918-921, 2001). Nilsson dent assertion that cephalochordates quired by Darwin's theory are nev- and Pelger's imaginary population have "primitive eyes" is simply un- ertheless "plainly implied" through- ranges over space and time in a way true. out Nilsson and Pelger's paper owes that could not possibly be disci- Now that I have swept away a few nothing to the facts and little to plined by the data. straw men of my own, let us see common sense. If changes in their what is left to clean up. In my essay model were really random, their NICK MATZKE believes that Nils- I wrote that Nilsson and Pelger temporal estimates would be apt to son and Pelger provide a mathe- made no attempt to discuss the cost- change by orders of magnitude, a matical model for the development benefit payoffs associated with an point I made in my essay and again of the eye. Let us be honest: beyond improvement in visual acuity. My in my reply to Dan-E. Nilsson a few finger-counting exercises, aim in ,discussing the reconstruction above. In my essay I also questioned there is no mathematics in their of the fish skull was not to argue that Nilsson and Pelger's decision to model, and while their references eyes came first or that bones did. hold selection pressure constant do contain some legitimate mathe- Paired sensory organs and bones are over time. In this, I found myself matics (nothing beyond second- characteristics of the vertebrates. echoing John Gillespie (The Causes semester calculus, but also nothing Plainly they evolved together. Plain- of Molecular Evolution, 1991, p. 294). to sneeze at), their references, as I ly, too, one function of the bony "[W]e must be concerned," Gille- have shown in patient detail, do not skull in vertebrates is to provide pro- spie writes, "with models of selec- support their theory. The task of tection for the paired sensory organs tion in variable environments. How modeling the eye's complicated ge- located on their heads. The protec- could it be otherwise? Natural se- ometry from light-sensitive cell to tion racket, as every Mafia boss is lection is a force adapting species to fully functioning eye is utterly and aware, does not come cheap; but their environments. Environments completely beyond our powers, as a Nilsson and Pelger, in adding up the are in a constant state of flux; selec- glance at any textbook dealing with benefits of visual acuity, did not ever tion coefficients must be in a con- embryology would show. bother to consider the vigorish. This stant state of flux as well." What is Mt. Matzke devotes the greater is such an unobjectionable point that good enough for Gillespie is good part of his otherwise interesting let- I cannot imagine why Mr. Matzke enough for me. ter to doing battle with various "cre- found it fishy. In approving of the value chosen ationist straw men." It is useful I very much appreciate the letters by Nilsson and Pelger for selection work, I am sure, the more so since from David Safir and Norman pressure, Mr. Rosenhouse writes the creationists are never named. Gentieu.

[26]